This action might not be possible to undo. Are you sure you want to continue?
https://www.scribd.com/doc/46264541/signalsandsystems978047198800730478
08/14/2013
text
original
This Page Intentionally Left Blank
risbane * Singapore * Toronto
Copynghlg? 2001 hy John Wilcy iy Suns I.:d Baffiiis Latic. Chichester. West SUSSCX,POl9 lIJn, hgliuid
N.iiionai 01243 779711 Inlciiralional (+44) I243 779171
eInail ifor orders and customer semicc znqiiirresi: c~booksiiLu,iiey.co.uk
Visit ous Hornz Page oil 1ittn:ll
~~t~~:i!u;vi\u.~~iicy corn
All Rights I<escrveil. No pan of this publication may bc reprodiiccd, storcd iii a ~ e t ~ ~ csystem. or traiismrttcd. iii any val form or by ariy means, elecWmc, mc or otlierwisc. escepi uiider tlic terms or the Copyi?ght ncsigus and Patcllts ued by the Copyright L,iccrising Agency. 90 Tottmhun Court Rwiid, 1 exceptloo ofruiy niateiliii supplied sp exclusive use by i l i e purtliaser of the publiciiiion. Nei1hc.r lhc authors nor John Wilcy & Sons Lrd acccpi any respon&nlity or liahiiiry Sor loss or damage ~ C C R S I O ~ C C ~ io any person or pmpeify through using the material. it1StriictlOiis, inclhods or ideas coiitnined Ilereiii. or acting ix rcti.;iinmgfrom ocrmg aa a resuit of web uio. The authors and Publisher expressly dircilmi ail niipiicd urarrariires.
this work.
WileyVCM Vcrlag Cimbli I'apprpclnllee 3, D69469 Weinherm. Germany
John Wley & Soos (Aoslmlia) Lid. 33 Park Iload. Milton, <&xiisland 4064. Australia John Wiley d Sons (Canada) Ltd. 22 WurccsWr Rurd Re~dale. Oiltarin. M9W iL1, Canada
l,il?mrp qf Cmgi,css C ~ ~ / a I o ~ i ~ ~ ~  i n ~ F u hDutu t i o r r liLn
Gimd. Benid
'
Rnhcratein. Alaxanrlei. Stengcr
p. ctn.
Includes ivhliographical Efcrences &nd iudex E R N 0 471 iikXOO 6 I , Systenr theoiy. I. Rnhcnstein. Iludolf. [I. Stciigcr, Alexmdcr. Ill Title.
SC?lCS
Q2YS. 0 5 7 2001 003 dc2 1
~
200102434!,
British Librag, Cafalugrringin ~ ' a b l i ~ u ~ Data iiin
h calnlogiie record for this book IS avaiiable Liuni rhc Rrilmh Librag
I S M 0 d71 98800 6
Produced fiox I.aTeX tiles ylippltcd by the traildelor Printed and bmind in Circat Wnraiii by Antmy R(iw 'Ih i s bomk ra printcd on acidfree p p e r reiponsihly for tvhich at lens1 two trees are planted for each one wed for papm production.
reface
1 I n 1.1 1.2 I .3 1.3 ~ r ~ ~ u c t ~ ~ ~ i Sigrrals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Systcm:, . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OvPrvicw of the Book . . . . . . . . . . . . . . . . . . . . . . . . Exerciscs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xv
1. 1 5 13 13
Is of Continuous LTISystems 2.1 DifferenLia. Eqi ions . . . . . . . . . . . . . . . . . . . . . . . . . l 2.2 Block Diagrmis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 StateSpace Description of LTISystcms . . . . . . . . . . . . . . . 2.4 Higherorcler Differentia.L Equations. Rloclc Diagrams and the Si ivlodel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Equivdcnl. StiLt.6?SI’ii(.e. R.epresoiit,atioris . . . . . . . . . . . . . . . 2.6 C‘ontrolla. arid C)lnserva.ble Systems . . . . . . . . . . . . . . . . . hle 2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~  ~ y sin the~ ~ ~ u e r i c y  ~ o ~ ~ ~ i ~ i t e ~ e 3 .I Coniplex I+equt:rrcies . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Eigenfunctions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Kxercism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4 Laplace ~ ~ ~ s ~ o r ~ 4. L ‘l’he Eigenhtiuictiorr Fornuilatiori . . . . . . . . . . . . . 4.2 Definition of 1. he Laplac(: ‘1i.ansforrn . . . . . . . . . 4.3 Unilateral aiitl Bilntcral Lit.place Transforms . . . . . 4.4 E.xamples of Lapliice ‘lYarisforms . . . . . . . . . . . 4.5 kgicin O CollVt?J‘g(Tlceof the! LC?l>liLC ’&msforrn . . f t. 4.6 Existence and linic~ucnessof thc Lapla.ce Tra. iisform
17
17
19 29
61
....... . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . .
. . . . . . . .
61 61 63 64 66 72
1.7 Propcrtic\s of the Laplnce Transform . . . . . . . . . . . . . . . . . 75 4.8 Exertisw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
5.1 5.2 5.. j 5.4 5.5 5.6
lex Analysis and the Pnversc? Laplace Transform 87 Path Iiitc?gra. in tsheCorllylrx Pliuie . . . . . . . . . . . . . . . . . 87’ Is The Main Principle of Complex Aiialysis . . . . . . . . . . . . . . . 81 1 Circular Iiitegrals that. F,nclose Siugiilarities . . . . . . . . . . . . . 89 Ca.uchy Irit.egrids . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 lnvcrse Lap1;tce ‘1Y;l.asforrrt . . . . . . . . . . . . . . . . . . . . . . . 95 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
Analysis o Continuousf ISystems with the La ~ a ~ $ f o r ~ ~ 6.1 Syst.ern R.~sponsc Bilatcral Iriptit. Signals . . . . . . . . . . . . . to 6.2 Finding ihe Syst. Fmx%ion . . . . . . . . . . . . . . . . . . . . . em 6.3 Poles ;ind Xcros of tlie Systelrr Functioii . . . . . . . . . . . . . . . em Ecpat.ions . . . 6.4 Detcwrtiuing the Syst. Funct.ioii from Ilif€ererit3iad 6.5 Surnrrin.risiiigExample . . . . . . . . . . . . . . . . . . . . . . . . . 6.(i Combining Simple LTISystems . . . . . . . . . . . . . . . . . . . . 6.7 Comhinirtg LXISystems with Multiple Iirput.s arid Output.s . . . . 6.8 Arialysis of St.nteSpacc Descriplioris . . . . . . . . . . . . . . . . . 6.9 Exercisw . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
105
107
110 112 113 115 118
121
120
nitial C o ~ i ~ i t i Problcrns with the aplace Transform 12 oi~ 7.1 Firsts.Ortler. Systc?ms . . . . . . . . . . . . . . . . . . . . . . . . . . 125 7.2 SecondOrdcr Systems . . . . . . . . . . . . . . . . . . . . . . . . . 135 7.3 IIiglierOrdcr Sy ms . . . . . . . . . . . . . . . . . . . . . . . . . 137 nt of the Proct:thires for Solving Initial Condition Problenis 1.18 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.19
~ ~ o ~ i v o l ~ t i o r Impulse and i 153 153 8.1 Motivttt.ion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Tirne Uc!liriviviour of ;in RCXircuit . . . . . . . . . . . . . . . . . . . 154 8.3 ‘I’he Delta tmpulse . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4 Convolution 16’7 8.5 Applicntions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187 8.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
The Fourier Transfarm 195 9.1 Review of the Lap1 Tra.nslorm . . . . . . . . . . . . . . . . . . . :1.95 9.2 Definit.ion of tlic Fciuriclr Transform . . . . . . . . . . . . . . . . . . 196 tj.3 Similaait. wid Differcnccs hntwccn Fonrit.r a.iltZ L ~ ~ > l i i (‘Ika.risf(mns 198 iw X! . . . . . . . . . . . . . . . . . . 200 9.4 Examples of the Fouricr Transform
['ontents
vii 208 213
7 1
9.5 Syininctries oi' the Fourier Trtmsform . . . . . . . . . . . . . . . . . Fourier 73rm:ifc)rm . . . . . . . . . . . . . . . . . . . . . . . 9.7 Yrop(!rf.i(?s of the Fouricr Trrziisforrri . . . . . . . . . . . . . . . . . 9.8 Pimcval's Theorerii . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.9 Corrc3lzLtion of' Deterrniuist.ic Signals . . . . . . . . . . . . . . . . . 9.1.0 l'irucWnndwicl(,h Product . . . . . . . . . . . . . . . . . . . . . . . 9.11 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PO Bode Plots 10.1 Introtfticiioxi . . . . . . . . . . . . . . . . . . . 10.2 Chr&ribution of' Iu&ivi(hd Polcs mid Zeros 10.3 Rode Plots for Multiple I'oles a r i d Zeros . . 10.4 R d c s for Bode Plots . . . . . . . . . . . . . . 10.5 Compltx Pairs of 'Poles t ~ . n t l%(>rt):i . . . . . .
4
225 227 232 236
2411
. . . . . . . . . . . .
241
. . . . . . . . . . . . . 2/12
2/16 248 2/19 255
. . . . . . . . . . . . . ............ . . . . . . . . . . . . . ................................
1 Sampling and Perie, 1 261 11 1 Irit roduetion . . . . . . . . . . . . . . . . . . . . . . . 261 11.2 Delta lrrryulse Trttiri arid Periodic F'uiictioiis . . . . . . . . . . . . . 262 11.3 Sernpliiig . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 11.4 Exer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2bG
295
. . . . . . . . . . . . . . . . . . . 295 12.2 Sorrre Siniplc Sequences . . . . . . . . . . . . . . . . . . . . . . . . 297
3 2.3 Discrc+te'l'inic Fourier Traxisfomi
. . . . . . . . . . . . . . . . . . . 301
:$OH 308
12.4 Ya.mpling Continuous Siqinls . . . . . . . . . . . . . . . . . . . . . 12.5 Properlies of tlic . A F Traiisforiii . . . . . . . . . . . . . . . . . . . . 12.6 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Exmplcs . . . . . . . . . . . . . . . 13.2 Hcgion of' Corivmgrrice of the zTransforim . . . . 13.3 R.c.lzttionships 10 Ot,lic.l, 'I 'x.;iiisfc)rmalitjiis . . . . . 13.4 Theorcrtis of Lhe 2 Trttrisform . . . . . . . . . . . . Tran&)rrn . . . . . . . . . . . . . . . . . Diagrams in t.Iw zPlaiie . . . . . . . .
312 315
......... 315 . . . . . . . . . . 320
. . . . . . . . . . . . . ........ . . . . . . . . . . . . . . . . .
322 326 328 332 334
339 339
339 340
................................
iscreteTime ~ ~ ~ ~ ~ 14.1 Iril.rotliiction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2 Lincnritv arid 'I'imcbinvariance . . . . . . . . . . . . . . . . . . . 14.3 Limw Diikitwe Eqmiictns ufith CoristntitcC'ocficients . . . . . . .
y
Vlll
...
Contents
14.4 Charactrristic Seyuerices iilld Svsteni Functions of Discxcte LTTSysteins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.5 I3loc.lr 13iagrams Z L I IStateSpacc . . . . . . . . . . ~ ....... 14.6 Discrete Convohxtion and Impulse Rwponse . . . . . . . . . . . . . 14.7 Excrciscs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
15 Causality arid the Hilbert Trarisfovin
341 346 350
362
15.1 15.2 15.3 15.4
307 Causal Syytems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367 C‘a~isal Yigirnls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 370 Signals wilh a OneSirlcd Spec;trttrri . . . . . . . . . . . . . . . . . . 374 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 378
16 Stability and Feedback Systems 383 16.1 BIBO. Impulse Response ancl Fkcyiieiicy R e s p ~ s Cimw . . . . . 383 t~ 16.2 Causal Stable LTISystems . . . . . . . . . . . . . . . . . . . . . . 388 16.3 FwdhiLck Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 394 16.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
17 Describing Random Signals 403 17.1 Inlrotiuction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303 17.2 Ij;xpect?ed Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . 405 17.3 Sta(ionltry Random Processes . . . . . . . . . . . . . . . . . . . . . 41.1 17.4 Correlat.ion Functioiis . . . . . . . . . . . . . . . . . . . . . . . . . 416 17.5 Power Density Spectra . . . . . . . . . . . . . . . . . . . . . . . . . 425 17.6 1)escribing Discrete I<. antloin Signals . . . . . . . . . . . . . . . . . 4.30 17.7 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 432
18 Random Signals and LTISystems 437 18.1 Conhining Random Signals . . . . . . . . . . . . . . . . . . . . . . 437 18.2 Response of UrlSystems to Random Signals . . . . . . . . . . . . 441 18.3 Signal Estimal ion IJsiiig the Wirucr Filtcr . . . . . . . . . . . . . . 352 18.4 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
Appendix A Solutions to the Exorcises
465
Appendix B Tables of Transformations 563 13.1 I3ilater;tI I, ap1ac.e ‘IYiIrlnfornl E’nirs . . . . . . . . . . . . . . . . . . . 5631 13.2 Properties of the Bilateral Laplace Transfornl . . . . . . . . . . . . 562 13.3 Fouriei ‘I’raiisforni Pairs . . . . . . . . . . . . . . . . . . . . . . . . 563 13.4 Properiirs of the 1~‘oiuirr ‘I’raiisfcrrm . . . . . . . . . . . . . . . . . 564 B.5 ‘Twosided z Transform Pairs . . . . . . . . . . . . . . . . . . . . . 565 €3.6 Propc‘rtics of the zTrmsforni . . . . . . . . . . . . . . . . . . . . . 566 B.7 Discr cteTinie Fourier TPmnsformPairs . . . . . . . . . . . . . . . . 567 B.8 Properties of tlie DiscreLeTimc> Fcruricr ’I’r:uisfor.m . . . . . . . . . 568
Conteuts
ix
This Page Intentionally Left Blank
as the text progresses. engirireriiig tlcnlilntis a ins tlirorv logically begins w i t h the An introduction to tion: linear. explanatory steps. subtfistiplincs. this book as an easily ilccrssiljlct introduct iori to rical rngiiwcring.ools is extxaordinarily important for engineers.xamples then illustrate further aspects of an idea.lie siibjcct iitifortunately can prow ciifficiilt. and their theory has attaiiied advanced maturity md elegance. The content itself is nothing I has already btvn dcmibed in o h r books. il the rcynired tmd drhrrvrd neadtmlic progress does riot Inaterialise. [. i stmting with ail example and thcn generalising from it. we con1. since failure to Icwn the fundamentals of systems theory oiroultl 1i:we catastrophic reper(*iiss”iow niariy 5iibsPquont sul:. To aid untler81. By meam of small. we generally use i ~ tinductive appro~wh. systems theory is a part of the core curriciilinn of modern clcctrical engineering and ser assthe foimdatian of a hrge niimhcr ol. . Wherever a piclure or a figure can enrich t h e text. and.t inight right unpopular. we provide onc. For stiiderits who are confronted with the theoi y of lincw. access to special cry of systeriis fh(JIy.iiiuously order the statrments of systcwis theory in their overall contrxt. we nwcr ncy. Accor(liugly. What is new is how we deliwr the material. Iiid~rd. 9:tturu rlv the retder citn ss whet her we have achievcvl our goal.i tical ioirririla or a tlicorem 1 altc+ pr~~ceclence over its proof.Analysing arid tlesigriing systcms with the help of suitable malht?niatic:al i. Accordingly. tlir stutlriit must persewrr.lect a discussion of its appii(*atioris and consequeritcs! The iiuineroiis exercises at thc end of each chapter (with detailed solutions in the appendix) help to reinftorce the reader’s knowlcdgc. iii this book it rliwiwion of thc irnportancc of a uiaLhriri. Furtlierniow. ‘lhis cot1ltl he diie to the abstract natriirr of the sii coupled with the deductive arid u n c k r presentation in some lectures. tirncirivariant systems.sriding. jc far . W find applications of such systems eve erywhere. we aim to present the abstract conccpts and in(erconr ns of systems tlreor o simply as to make learning easy m t l fun. tirncinvariant syst rius for tlric fii st time. While we might omit the tlcrivation of an eqiiation. the su1)jec. IIowevw. Additional c.
Giidde. large portion of text. tinieinwriant syst. who pxodiiced niuny figures. we discuss random signals. A'Iarioxi Our student. the rnt31mhl in this book can be worked through complet ely i n about 50 hoixrs of lecture8 and 25 hours of exercises. we do not assume familiarity with thern. An engineering curriculum often encompasses complex funct ion 1heoi. we are convinced that it will also be nscfnl for practitioners. the university.ioIis becomcs evident o n l ~ through the characteristics of lincar. Finally iiiinie~oiiswaders of the Cerrnaii wrsion r cpnrl.xii Prefuce Although we have written this book prirriarily for students. Finally.roduce timcdiscrete signals arid svstenis iLrld so cxtorid (. An engineer who wants to brush up quickly on some subj will apprvciatc the easy rcadability of this text. This book is altio siiitablc for selfstudy. wF treat sigiials and s y s t e m in parallel The purpose add of dcscribiiig sigiials by meaiis of their Lup1tix:e or Pouricr Lr. Stcfan von der Mark a i d Ilubert Itubcnbauer derrionstraled trernvndous commitment in typesettiiig and correcting the book as well as the solutions to the rcises. This book evolved out of a course on systems theory arid the corresponding laboratory excwises at the F1ieclric. and its rnaiiy examples. T n our presentation we ernphabisc tkic clcar concept of Rigtm functions. Assimiing fulltime.t>ms. whose form i not s changed by systems.11Alexander University in Erlcuzgeri~~rriherg. conwntrated work. All formulas ilritl most of the figures were typeset in LaTeX and then transferred onto overhead slidox I W W itre most g r a ~ e eful that mere used h r two y e t m in the? nte criticism Inelped us to to some 200 registered students wh debug the presentation and the typcset equations. who typed and corrected a. we int.. T n addition. . Contrary to some other books that first introduce detailed forms of description for signals and only rnuch l a t c ~ systems. prt ntation. its practiceorient etl. We thank Ingrid Biirtsch.q)h~m Schahert.liircnr algebra) and basic luio&cdge 01electrical circuits. which elegantly allow us to couple XI external and an interrial cornponcnt of the systcw response. After covering sampling. the material is also suitable for iise in the third or fourth semester.oncopts fariiiliar from thc conlinuous c asc" r T l ~ r w aftcr discrete and contimioils sigials and systems are treated together. Sl. wti liaridlcd the layout of the book ourselves a(. A s such.liec. Our presentation bcgins with continuous signals arid systems. one yea's students read thc first version of the manuscript and suggested diverse irnprovernents.y and probability theory as well: dt hough thesc fields are lrelpfid. The course is compulsory for students of electrical engineering iii 1he fifth semester. the materiiil can be covned in four to six weeks. we use state space descriptions. Assuming that. which tire very important today. R S well as Susi Kosdiiiy. t?hismathematical knowledge hi^ been ucquired earlier.zrisforxnat. To take into account initial staks.ed typographic erioib arid sent comments by email. assistimts Lutz and Alexander Larnpe. ?Ve do assume knowledge of the f h dumerit als of crigiricering iiiathornatics (diEerentia1 anti iiitcgral ral~ulub. To avoid the arduous m d seldom perfect step of correctiiig cameraready copy.
Bcrnd Girocl Rri d olf RalJenstein Alexaader Stcnger .
This Page Intentionally Left Blank .
I I I i @ = 100 200 300 400 time (ms) .+0 I I I I .
hp wall.1 0.2 1.2: Temperature curvv for n hoiw nali 2500 i Jan 5. 10 8 6 4 2 0 a 0.06 0.1996 . Zntrodiictioii would hme ic different temperature c u ~ v c .04 0.irne hew is.08 0.14 location [m] FigitrP 1.12 0.02 0. 1996 June 28. t. ti pnrrtrrwter of a family of ciirvcs: thc indcpcndcnt coiitiiiuous variable is the location in t. Iri contrast to Figiire 1.1.
4:Frequency of e m w d marks for a test in systems theory The signals u7e have ronsidcrcd thus far have been ryiiant.6): two coordinates arid thnc.at represenls .7 3. The dependent variable s(x.d according tJo various rritmia. wc h:me c2 depcntleiicy on t k r w iiidrpendcrit vruid~les(Pignrtt 1.0 1.s on two or more variables. stotk index.iori to pict. %'c citll these two. $ 15 10 Y L 0.1.iticas that clcpend on a m q l e irrtlepcdcmt variable. Here both axes represent independent variables. is n greyscale value between the extremp values black wid white. In this booh we define it signal BS follou\s: L _ I _ _ _ _ __ Definition I: Signal _______ A signal as a fil. how eve^.ures.7 5. The grcyscalos of Figure 1. Simals 3 averngr stock index) arc ~ v l ~ onumbers and so likewise discrete.3 2.0 4. these are continuous signals.1. I I . 4gaaJs caii a ~ s i t ~ m ctifftwnt iorms.5 rlepeird on both thc .znurnit c*rrtaininformation.01.3 1.3 4. 5 0 1.0 3.ires th.7 4. All (mr examples have shown parameters (voltage.7 2.nctaon o r sequence of val. but.E and the y coordinates. When greyscaln values cliaiige continuously over space or over space and time. greyscale) that change in rda tion to values of the iridependent variahles.0 2.thrccdirnenyional (or gcnerallp multidiineiisio~~a1) signals. Signals can l x c*l:wsifit. Z. tliezp arc yuantitics with t~t?i)c?ntlciic. 'L'lierc~by thc?y tr. ding exaxxiples hil\rp shown that. 1lie inost important o f which %resnmunarimi in Table L. Wlieii W P adcl mot.miperatiirc.1. frequencies.~ri$!orwcutzon.y) is entered along one axis.it.3 3.0 mark Figure 1. Iri this case bolh lc the independent arid tbc dependent variables arc: discrcte.
.
.
I X . sigii rstshlihhcs n relation&ip amoirg the si gure 1.ts sllrtl a sp I _ _ X .In tliis gcricral form we cair iriixgirie cl with t tie outside world via varioui.7 tlep1c.
. Ir respoiitl. allrrvvs systeiris theory a. h. physics. ‘I’hi. the system iinposcs bet wtwi its sigonrrits. This throry also prows suitable for describiiig I ~ I tcms Iliat caii bc Iiiitw small signal aiiiplitiitlcs bysi eins theory tiiiicin\iariarrt syst froin tlie practical pirjt)l(ws ol electrical rnghwring over i~iorc thaii a t c . 1 1 t heorv clocs not ciiconipass the implerrretit at i w ot a 1s relatiorirhips Clrat. elrgatit arid clear. econoinics.9. kit us considcr the systcni iri Figiirc 1.cai ning econorriy c~isiir:. ‘I his repicsmts tl assical core ciomain of s y s k r m t Iicw J.1. Clarity rcsults in bcmnsc separating the ctrt>iilproblems froin the gciieral relat. Aiinlysis a i d dcisign of clcctrical circuits e  Digital sigml processing Clorrrmuriiw t ion? To tlefiire thr tcrm liucazfty. ~ r r i o u n t of unclcariiess tliat enciurihri~s initial leavriirig iir systeiiis tliror y. timeinvariarit of systems.. Important a p p l i c a h n dolX*iLiri\ h r thr thcory of liiieai. to an iriput signal . unilorm rvprescJrital ioir of procc on tlonieiiis (P g. ear. 1. However. Iln important sihfirld of 4 ms thcor! is t Ire 1 l i ~ o r v liiicar. timeinvaria 1 cleclxicitl erigirr irig totlav inc hide. enginecrinp. systcllx ~ 7 1. biology) arid iisciyliriarv view..nncl i h cvell develtq~xl. r t ( t ) with tlic output signal $11( t )ant1 t o the input signal .with i nals.. Svstems lheory I Pscrit s a powrrfid xnathcwiaticaJ tool for tlw stud\ and mentation details h l p s to inniritaiii bccttusc ontitting the i c~ebigrr or s t b c ~ r y focus is 011 the formal thc an overview of the ovrrall systcixi.. In s ny sprcialisation for specific applinat me of the irit erc or111 ions. this i s countcirtvl by t h c drawback ot a cer l a i r 1 .cZ(1) .k n n u l d ~ d griirial form. priiiriplc. The high t l c y g w ot abstraction b gs the advcirrtages of learriiiig ccwioniy m d clar its.2.iorisliips i s clcvatcd to n.
(1.In gcrieval wc canriot inakc this step>but fbr inany relationship bet.a\y.9: Definitiori of t~ linear system ( & I1 are arbitrary comnplc?x constants) 3 .werii charge and voltage in a capacitor. I 1  . (:an be defined more generally a follows: s on of mput szgnals always the mdimdual outpzlt sagnde.r s y s t m s . we also use a more tangible term: ____ Definition 4: Linear systems Systern.s fw whrch the supe~ipostlmi pr. It.itiwn iripul arid output parmwters. from ( I.aple.mr:zplc upplics arc called 1inea.1). and l ~ i ~ e ~ n force aiid stretching of a spring according tu Hook's Law. The relat.ionship expressed in (1.he supcper~o.srtzon yrrrt.2) is called t. rc4atioriship between currcnt a i d voltagc on a iwistor given bv Ohm's r. I linear i Figure 1. bet.I) [lie output sigiid follows as F:xannpleu iiiclrtde kh(. tiicn _ l l " r Due to the grwbt importaxice of su(b systems.
9 tlie superposition principle is This cir6iiitaionvan be geiicralised for systmis with multiple inpiits and outputs .1.2. Syslenis 9 FOIthe s y s t m i iri Yigurc X .
is already mentioned. systems that are both linear l arid tiirieiiivarirmt play a przrtkularly important rolc in systems tAieorq. 1 . need not be linear. 2”1rrl.imcinvariaiii.2. The laws ol signals. However. we crnploy the branching circuit in Figure 1.11. voiw signal in Figiir establishes relationship betwccn two signals aiid is t h s il :>y the electrical nature of the inner workings and the enclosed coniporlents regial:irities.4.i i ~ ~ ~ ~ i i ~ i tt. the assipmerit of inpiit signals to output s arid rripacitors) allow circuii. The timedependent voltages u l ( t )and u ~ ( t ) zq)resciit contirnwus signals.1 Electrical Circuits A i an exunplc of the description of’ RTI electrical circiiit iis a wstem. i ~ i i ( ~ . we I I ~ O VLO tJir represent >ition : i. Ini roductioxi [I!)]. les of Systems 12.uci.f~. Tn The clirzracteristicsof LTIsystems and the tools for their analysis arc the sitbjects of subsct~rlcntchapters.ion on ttrr origin of thcse random. 6: LTIsystem A system that zs both tLmerrcvarscmt and linear 2s termed u n LTIsystern Lancu7.rinnt sgs.3 LTISystems Linear m s iii c iiot gciierally t . A s long as we have rio furtlicr inforrna1.10 could assuirit~ positive or negative values. systrrrls Likcwisc.In thr definition me could emhaiigc ii drlaycd aiid a noiidrlay~d signal and recognise that the delay in Figure 1.8.s ( i t l e d r us to reprevenl this circuit as a linear and timeinvariant system (LTIsystem).m input/outpzlt systcm ~ ~ 1.3. iheory for t he idealised coniporicwl.10 1. .tern). siicli n s the. l’lley have been assiguetl the acronym LTI. .
ht. This .11: E1ectric:el circiiil.1.igiire 1.12: Examples ol systeriia The speed of it car dcpciids on the poriliorrs of the gas and brakc prclnls.2.s a system position of gas pedal position of brake pedal speed acoustic pressure at the ear of the listener gross national product int~~es~ inflation rate unemployment rate 1. Svsttiris I1 system fJigure 1.
Irrtroducliorr .I2 I .
Thus ( oritinnoas oi disor digital.aiisa1 sy with rcv1. some XIS rcsult in noncausal scmw cases it is ea51er t o work wii Ii idcdisetl.ioniritlc~~ciiclriit signal (e.rin. unitlimciisiond or mull items arc systems tliat establish irlatioriships betwwn sigrtals with sticr.tiic~.oriil. ised A nuniber of thestb c.g. ms.2. By c uril r ast. this i a txaiislationirn a r i a i t i . we can classity svstwis sccorcliiig to various critc. iii ‘I’ablr I .ritcwa w c familiar from signals. In niofe g:cm\ral ontjirig to time irzsmia loca~.lexv.2: Criteria for classifying system?   discretc dig<titl cony~lexvdlled multidirncii8ional st uchastic noncastsal rneniorylws nonliriear timeva riant trmslat. 11 system is ( ailsal if ifs response to the arri of 2 I irrie signal does iiot hcgiii beforr t This sounds txivial. Furl herrrrorr.rit AS with signal\. of 11s the values of input signds of othcr times also piay ii iole. A digital system i s tlius onr that processes digital signal$. Thus the t wiporal clclay ion. realvaliicd o r c. signd at c% cc’rtaiii on t2ic valiic~ the inpiit signal at the hame t iiw.~liie(~. For rneniorylcw systems. id mernory hvstrrns these other valiies rriiist be ihose of previ i.3. noiit. caussl ones. for i i by Lhe lmrs of rrittrirc ily (miss1 However.ionvaria. Overview of the Book 13 Table 1. there :ir eiris wliost indepeiidr is riot time. ‘l’hr most important critexia are smi~xiay. the r SP to a tiuic.1.
Exercise 1. In order to be ablc to describe the impulse response mathematica.mge (Chapter 3 ) 1ea. In the subsequent chapters continuous axid discrete systems and signal: arc treated in combination. this occurs primwily via system description in Another kind of characterisation of LTIsystems i the impulsc rcsponse is discussed in Chapter 8. and Chapter 16 prcscnts st.. arid in Chapter 14 we use it to mialyse discretetime LTIsystems..he Laplace transformation.ability characteristics ol systerns.ransforai and its cliasa ris&.mt coefficients.mpledand periodic signals a wcll as the sampling theos rem and leads u s to discrotetime signals and their Fourier spectrum (Chpter 12). Studying IXlsystems over a frequency r.i(jn via.ms is the snhj Chapter 1.ds1 s to the reprc1 sentation of continiioustime signals with the help of the Laplac:(: transform. in a. is the Fourier transform.:I concerns sa. the discrete counterpart of t. we introduce generalised functions in this context.equations with const. Although linear differential equations with consta.ddit.1 Are the fdlowirig sign& . The graphi.tion is the subject of Chapter 15. Finally. whi.lly.sponse of systems by means of Bode cliagra. cli scrot et irnc itnd / or digit d! ' II aj nitmber of days of rain per inonth b) average high tcinpcrature per month c) current tcmpcraturc .ioii. The characteristics of causal systems and signals and their description with the Ifilbert transforma.egral transforniation cqual in iniporttialce to thc Laplace tramform.ch svc discuss in detail in Chaptrer 4.nt coefficients arid specified start. discuss a frequency response representation of rniidom we signals via power derisily spectra.cal analysis of the Gequericy rt. whose characteristics and laws are discussed in Chaptcr 9.erns with the Laplace t. In Chaptcr 13 we handle discrete signals with thc xtramform. sralues are no longer LT1systems.i mpl i t iiti+cli scwtc.the Laplace transform and its fundamentals in complex function theory. thc systc~ii tiinc%ioii is the subject of Chapter 6. Chapter 18 is dcclicltted to the yuestiori of how expected vdues and power density spectra are. The analysis of LTIsyst. LTI methods can be elegantky exteaded for this important class of prohlems (Chapter 7). In Chapter 17 we introduce random signals and their description via expectcd \dues. An inr. modified by LTIsystems. Chapter 5 presents the inverse forrnultz of .
amplitudediscrete. l i r i r w . ciiusal. analoguc..h mcrnory. An ideal A/D converter could be constrircted as follows: a) Which of the signals 2 1 . timeinvariant .I . Exercises 15 xercive 1 3 . memory. ZUP digital? 1)) For hot 1 systems. discretetime.n::3 RSC analogue. sprcify whether they 1 wit.4 Which of the following system descriptions designate linear. . . or carnal systems? a) y(t) = X ( t ) h) y ( t ) = $ ( f ) c) y ( t ) = x ( t .4. Exercise 1. tinicinv:~rinnt. T > 0 (delay component) T > 0 (accelerator) d) y(l) = x(t A 5”).2’).
signal yl (t). c) redvalued.ws(ojt) [l+ r r r .T ) produces die ou(put 5ignal y:$(t) f t  T).. Are the systems a) linear.zkt>ijn unambiguous statement about the above systerri chararteristic*s? 1 Icfcnd your ariswer.r(t I . ~ ( ( t. . T ( I ) arbitrary . d) mcnnoryless? Exercise 1. Inpiit of ~ ( t ) produces the outpiit signal y z ( f ) 9 (t .5 Two systems S and 5 2 respond to a inpiit sigrlsl x ( t ) with the ontput yigiials 1 n g1(1) = S l { c ( i ' ) ) ) y*(f) = S J { S ( ~ ) = rrr. ~ 2 ( . is tiiwinvariiuit mid liiiear. '% (Y mmsurcnients of. input of I C (~ t ) produces the outpilt.) ] ros(d1t ) 711 E IR .6 A l)lac~lcbo\isystem is to be examiiid 1 o detcrinine whether it. b) timeiiivariant.T ) 1 + wz(l l ( t ) = r l (1 . 2.1i f i) ! I ( [ ) = . Exercise 1. the system yic)lded the following data: I 1.Y') + . htroduction ?'(t)). can you Iri.r(t).
s .
'Ylic. Figwc 1. electrical c*mrc. itrid vi'~will thercfoic cmmiirie them in imm dcptli ~ .lrcvcr. Tliiis icmilt s iii ordinary difT(wntia1cqiiatioris coefficients. The resrtltiiig electrical networks (for cvmplc. foi esaniplc.g. 1k)logy or economics. lrvdr anlic ancl i. often liiirar enough to b3 . liowcvrr.11) CRTI be aiialyscd idartl met liods. for eminplc. capacitors lravc i t l ~ a l capacitance. lilic1 electrical circuits. mechanical. wires becorrie ideal c o i d i w t ois. t~lectrical voltage) ancl flow quantities (e. pneuinalk.hltwis.h(vnitil t o diffcrcntial <qiilrltioiis.siblc to igirore the spatial cxparisioii of clcctrical winpoiicnts on a circuit I)( d . unsuitabkl foi 14ir ficlcl of fluid tlynamics p.&'qnat c dellrcl by sirnplc coniponents like resi second step is t lie rcylacerncwt of tlic pliysical coinporioiits with their iclcd cqiiiwlents. Their witliin a n electrical circuit ale. u ~of great irnporl~nce. in which only t hc3 inpiit a i d output signals arid tlirir drr c u i also he applied t o othri pliysical aLraiigerwiits which. Iio\. they a . The sirirpliticatioris merilioriecl are of ('otirs~ always peiuii\siblP Ordinary iiot cliflcientisl rqim t ions arc. In niany othei UWS. real r Psis(aiices k c o i n c ohmir . can be clcscribcd by poirirtial (e.g. etc. TimeDomain hloclc+ of Coiitimous LTTSvstems In marry taws it is po.18 2. aiid instead work t h equivalent circuits coiisistirig of c OM erit I atc4 (+me Senicctnciuctor tievi alp aii exalriple o f tlliis because tlwir cornphcated irit eliaviour ( w i oiily h r arciirstelv niotlelled using solidstal c physim. same applies to other kinds of from cIreriiis(ry.. fo1 exarnplc.nt The anal) sis ac cortlingly siniplifiic~s ). mesh or nodal a n d y [ 18*221.
To show lincaritv IW consider the tn7o tliffererit iripu( signals X I ( t )tmtl . + Modelling of an LTIsystem iiidq)eiidrnt fi on1 i t h realisation. to hc cqunl to zero clifrerent hrrearly ~ridtyfnderit sciFoi a givcn fimction . without det ails of thc system’h ir it ei no1 betimiour .z) leads to the solut. RC? igriove possibly given initial conditions.3) Tli~ greatest index N of cl xronzero coefficieii( (1: R. cleterniirios wliat is isallrd the ordc~i o j f h e drflf rrntiul f p a t i o a .ates . . i j ( O ) . Lct 11% i m ’ show that (2. 3 ) .10.3) repremits it tirwi3ivari r T t cm Thl ough mbs l i t u t i o n of variables t’ = t . 3 ) verifies that y i ( t ) = Aylji. outpiit sign.pixt sigrial of ihc> system. Tliis mrthotl fiilfills OUT initial requiieniexits..T).2.z in ( 2 . but not all of tlrt.and thcretorc the uul.9 and 1. input outpiit relationship. . thcir iiifluence will l x (fisctiss~lin i irr Chaptci 7. corffic*icvitsNc. 1.For il p~iiticularsoliition.2.present inore irifbc)Imstion than cliilcrcntial equation show riot only thc input a i d outpiit signals h t also i n t t ~ r i a s t a t P s o a s l f only t tic inputoutput rclat iomliip i a of iritcwst t h i the choice ol int vxrial st. it f o l l o ~ iriiiiirtf s that r ( l .m of the.r(t) t’rieic are up to lutioiih g ( t ) to (2. Every system that citii be nioctelled. T3cpiel.3). c ( f ) is t hc input signal. 3 ) tlcstril )es a. This means we Imve foiintl our first method kor rrrodrlling such s w t pin5 in the folm ol a tlilftLient ial cqiic2tic)ii.anf coefficients (2. tising linear dillereritid rtpiations with coiist. continuoustime systeiir.x. wp 11ccd to gibe N conciiLiori5. Q(0). For now. ra (wi rc.3) is thus an LT1 system. and y(1) is t’ttc.d.ion y(f .t h e di ritial c~jnaticiii. The dif€aerrritial eqiiation ( 2 .) 1 L 3 ! / 2 ( ( ) . In order to chsra c w rcfm back to Ueiinitions 3 ant1 5 a i i d also Figs. if . we h>t A T = N aaid allow some. In order to simplify this discussioir. .a(t) a n d the corresporitlirrg solutions y1 ( t ) and pd(t). is a soltition of. Plugging llrelinear c~4ucttion r j ( l ) 1Ay1( t ) Bsdjl) into ( 2 . Block Dictarams I9 (2.ciitatit. FVI initial condition prohlcriis. these would be N initial coritlitioris y(0).
(2.4) .
fmir:t. ciitii ( 4 <irt)itruy.WO cw.iori. t h t i s not. urriving at diwct torirr I1 showu in Figure 2.: of I lie clifkwritial c ~ l i r a t i o u hfore important 1.iplwx cocfic~icutsof dirckct torm I1 eflicicntl. We can therciore unite t hr I. ctrily modelled by t ~ C J( oriPsporidiiig differeni. thc mi.2 run in pxrallel n i n w t h r t inpui sign& of the integritLors from tirnc t = 3c are eqi~al. anti i t is .3)but i s furllwr g i w i an iriteriini striicturc IIV clirrcl form 11. thiy form requires lllc only IV integrdtors.6.hout l l<nowledge o tlie tictlliil realisation of the systern.1 with the hclp of a frequencydoniairi model. A Y with direct torm I. liowever. t lie minimnnl nimbrl for ~LII IVordcr dif€(xtw(.3.3.cxnr2es.i. this C igiiniciit of stirtes is.tw.l forms. . z = 1.. Wit. The signals .u.di. Uc mill show this gonrral concept more elt:gi>titly iu Chapter G. Both cascades of integrators iii Figure 2. we will just view thit: property as a uscful rnsumnption. At the moment.ialrtqiiiltion. . . Black diagrams that usc thc miainium iiuniber of"cnergy stores (integrators) for the realisation of an Norder differential equation are also called canonzt.N at the iritegratcir outputs describe t h internal &te of a SJ st. tirid so are tJit3ir outpiif sigirak. ol c:oiirw.id t ~ p i aion (2.. ~ ~ Wr c'onsttructcd direct form I clireotly froin tlw ditfcrc~iitideyuatiori.
2: Dotjvatkrr of tlie block diagram for direct fobrm I1 from ilircc:t. To vctrify diiect form 11 the inpiit signal r and tJir oirtpit sigrtal y irr t.ly through the uppermost path.hes iiitenid sigizal q1. rclatiorisliips iv 1 (2.lwi signals me linked f luough ~ l i c states at tlclc integrator oul puts.22 2.vcl to consider this path us wcll.3) thr.erms of t. 'TiriieDornain Models of Coiitiiiuous LTISystcn~s Figurc 2. hrin 1 clcar that systcnis with this structure s a t i d j the difftJreiiLicdeqiiatioii (2. and rirtpliasisp {.lw s t a k s 'L h~ iiipiit i ~ r i t loiit.3). To simplify the iiota tiort we irttrotiiw anot. aiitl of rour'sc we Ira.hat i L does not repiesexit a state.7) . Wr n 1 I satiqfies the same diffcrcntial equation. but also dirrct. FIom the block diagraiii (Fignrr 2. impioveii assmnpt ion that tlie two \tagcc of direct forni I can Iw intcrchangcd to creatr direct form IT.
2.f Figixt.:3: An LTTsyhteni in direct form 11 (2.8) (2.9) .
as wc lrad 1iol)ed.u... Irit erchanging ilic ordcr of iiilegrai ion and summation yidtls T h e labt s w n ove re(ognise as T as in ( 2 .4 dirt 1 Pigure 2 . 3 .ystcm in tli I1 indccct hdtisfy the tliffwcmtirtl equatiori (2. xam Me will coiibtmict a block tliitgrHit1 for the eqiiiit ion 4.J j i .4) is satisfied This sliom that the input and output signals of R .3 r { %I. iiitcgrat ioii .a y = .24 1 2.3). Time~Donisin Xlodt:ls of Continuous LTISystcun~ times. 8 ) > hcmx>t h c intcgial equation (2. ( 2 12) tte thc direct form 1 ltlocli cliagrani shown in Pigurc 2.
13) (2.(2.16) (2.14) Irifegratioii arid clmnging iriclites y i d & (2.19} .
Figiirt: 2.24) .20) (2.22) (2. form IIX (2.23) The I)lock diagram in direct form I11 can be fcmied frcm the c y t m t i o i i h (2.21) (2.5: LTlsystern in direct.
6.aiicl i s sliowri in Figim 2. .
Symbolic rcprcwntation of an opamp Figiii e 2.7.8: (Dpamp n ith fccdhack circrtit .Figurc 2.
As miditioii a i d mu1tiplicatiuii . .iiit configuralions. terns raii be innpleinentcd as elect. Stntc4pacc Description of LTXSj sterns ~ 29 From t Iiat follows thc required iritcgral relationship between I (11 ( 1 j and ~ ( t ) : I.26) Tlic circuit givrii i s therefore an iiitcgrator of the inpu( volbage. ~ r o likcwiw posii hlc iisiiig opamps with othrr circ.rical circuits tlirecatly f r o m block didgranis. (2.2.3.
27) to (2.) L  U l ( t ) + Z(t) (2. this would rcsult in the loss of inforrnatiori aboiit thc intcrnitl rnergj7 btures that detcrrriiiip the s j r b t . we reprtwnt the equatioiis (2.l ( t .30) in matrix lurrrr: (2.& / 2 ( t ) = .30) n o n 1 these t trrw firstordcr dilrercntial eyuatioiis we ronlti eliiniruttc~ 12 and U I 11.28) (2. t ~ i n kxhaxiour.ial equation for the v c t ioris . (2. ancl olit aiii a thirdorder diffcrtlritial cquatioii in the form of (2.9: lLLC circuit The slim of the voltages in both mcshcs and tlir sum of thhc raruerrts in the 11odr y1eld d7 1 L% = R. Rowrver .29) The output voltagr yjf) is ttir voltage drop across the resisioi R2: )/(t) .32) are a sy re three difLereiit ial equw Lions ca aiid ail algcbraic equatiox a matrix tliffiwnt. Instead. TimeDomain Motlrls of Continuous LTISy&ms 0 0 Bigiirr 2.27) (2. coniwtetl fir stortlrr d ’ (2.32) of lhrer.30 2 .3). viewed ws P q u‘ .31) ztnd (2.
.3.34) tv hem x: a: y colunin vector AI input siglId5 cnlumn vector N state v i t r i a l h colurnn vector lip outpiit sigrtwls. St at eSpace Descriptioll of LTISystems 31 ( 2. U :) (2.2.
7) arid collecting the scdar variablrs gives 1lie state equation 0 I ..36) .hroiigli thc ?Ve haw already txxpr variables a r i d the input signal in (2. 04.32 2.. Siihtitut ing (2. TinieDomitin RIoctc~lsof Courtinnous LTISysteirrs d I hc input sigml of the first iritegrator t.7)... 0 0 . 0 0 + 1 :I: . 1 0 (2.35) in (2.
1 0 0 1 (2. ..42) .41) (2.....
110 rcstrictiorrs.31) and rnultiplJhg both values 2.:) .45) (2. Substitution into (.48) (2.49) (2.46) (2.gives t he W V I statespace equations ' (2.3! arid (2.47) (2." .
I I Figitre 2.11: The parallel forin of a systctn wilh rnultiplv inputs arid outputs .
(2.52) .
(2.56) .53) y 7 [1 22 (2.
.
in fact. s~\iric .sponcEing block diagram Equallv.at. t. as these iis in canonical form.ion of iritrriicrl statcil is c ~ ~ i r ~ iLSl arliitrsry 2~s :l v with the block diagianrs.ve11 lllc CoIlslrll('tlOl1 o f 1lrrs ordingly the pxallel block diagrams ancl the structurc of Illit' '1'11~ property of the wit tr 1hose 01 the clil'fi~rrntial as w l l 7'hi. be used di y to G r n i a c.o state iriodt+. The ~ i i i r i i t ~ o i i ' i cocifficicnth ol thcs l)c conihixrcd in the corrcct way in matrix form to givc tlrr si)c*all Allor. iiiatrix rnius matris (with suitable i w r i h ctifferriitial equation in the bottom row.orrc.* Observe how the states are numbered Cloyely linktd with i cliA'erentid1 equation is an equation o i orclcr M . wliilc all othcr matrix clrments arc cither 1 01 0. ever! drawn as a11 tqui\rtilent bloc4i tlliigrilili.tlie state model is firstorder dillkentin1 eqiiat ions. with a miiiirlral rtiiiiiber ~f states. T h e elerricnts of the st ~ T mat ric V can. With direct form LIJ. U'P rttsrict ourselves lierc. This m<ikea it clear that there must be soirie liiritl of I T. th.
2 .
4 .A system i s modellcd by t lie liiirwr diffkeiitial cqiixt ioii xercise 2.
.
8 .8. Esrrcises 33 xercise 2.2.
This Page Intentionally Left Blank .
.
tc.piue of oscillations per unit time j’: arid t. The traditiona.  (3. hlathumaticnl tools for the trailsition bctwceri tIic3 timetlorriain arid fieqiienrytlornain follow ill Chapters 4 :incl 6 (Lal~1.d ~ : ( t ) .c~ ‘1hnsform) arid in Chaptcr 9 (Four ier Transform).hc zero crossing relafive to time t = 0 .itly wci will ~ a k et gencral look at.lueiicy f is a r e d mimlx:r.Such a s i g d i s characterised by its a m l t d X mtl tlic fiumber . hut fir.doniaiii. The fret.1) .a.1~~ a sinusoitlal of f5igii. To Ber with the r c d pliasc p it’ dwxibes exactly tlie position of all zero crossings: z ( t ) = 2Sili(27rff 4 q ) .h Ix: cxamined iu detail in this and Llrc fdlowirig chaptws.1 ticfinition of frequency i s tlerived from real va1i. tlie concept of frequency arid the i ityroscntzition of sigrial5 in the rrerlueiic.
1.1.1: Exninpic.2: Heal itnd imaginary parts of the signal in Figiire 3. . Complex Freqnrricies ~~ 47 Figure 3. d a complrx cxponcurtial sigrral ' Figure 3.3.
Constant Rcal rxporicnt i:r 1 s i g 1*al Siriiisoitlal oscillation .
n{s 1= 0 iasler oscillation c/ / .
4 sliows tlrc close relatioiiship hctween llre input a r i d oiitpiit signal. Figure 3 4 Driving .I sybtern with cigrnlmctrori e ( t ) (3.Figure 3.9) .
Eigenfiiiictioris 51 or As an rxamplo tor malysis u s i u g coniplcx frrqueiirirs iwonaiit circuit iii Figure 3.10) .3.6 VCT consider the paral Icl Th(1 input sigiisl is ol tlie form (3.2.
4) (3. We put u(t) and i ( t ) into the differeutial equation (3.17) .1.For th(>crirrcrrt we thus m.ike t h c following st.vting poirit . ( 3 13) Tlie constants I 1 and I2 arc imltrmvii ancl nimt kw dctcmiincd froin th(' niodcl of the network.J l P i l l +I p t . r(t).
Eigcnfiinctions 53 .3.2.
15) and (3.16) for Ihc parallel rrsoriant cirtwt w e t ~ ~ l p .24) TIP impedance model is only appliriible to liiiear t iineiiuvaria systems) as only t h y havv c~xpoiic~iitial functions as cigcrifunctioi ~ A grirrral problem with derivation of the sy.pecifiridion of consider atioii of p2iysic. (3.20). (3. T&le 3.23) Thr oiitpiit m l i of tlir parallel iesoiiarrt circuit can also hr cxpr ~ form a i r ) = H(S.2 sl~osvs hr impedance fuiictions for the most iinportant cornpoiiriits.22) Froiii section 3.2 it follows that.2.21 ) ' where Z(S) =R i  S L.d imits of nieilsuremriil. function i s the cor WO fiiiidariiriital possiblitics: curisistent iise of units . Taking the cquatioiis (3. SL 27 1 + & R +SL t~ L C R 1 I.s"Cr (3. This rrlationship is called Ohrn's law in impedance tarrri. 1x1 response and the I. t Net woeks thal con of these cornponenta caii IN niodcllccl as an inipcdance.Pt'f + H ( .)U. and Irave: U P C t = Zjs) (3. 5 2 ) u 2 e 5 2 t . tlic systein fimction of the parallel rrsoriaiit circuit for the voltagc as input signal mid current as outpiit signal i s . i r ~ with (3.component impedance resistance c capacitance sC 1  inductivity sL with the impedance Z ( s ) .
The sjstem fiinrtioii of thc nctuwk shown in Fi"girrc 3.7 will l x derived. The voltage u I ( L ) i s tlir input signal.7: RC circuit . physical iritianing ancl t h a t errorc. b'igrne 3. rrz(L) thr outpiit bignaI. in calculation c m be detected due ( O iricori ly this rncthorf oftcn leads to unwicldy expressions.The advarit age of corrsislwt i r w of tinit5 15 that all values have a. We will piicy of Lriiit s m ith the following slmplr rxarriple.
itlid r compoucnt norrualisatiori for gerrrrsl refermw ~ a l u c s .hatall v d w s are ralculatetl using $1 units. which expre vitlues as a multiple of tlie corresponding reference valw.lie cxponcnt.iovi induce a correspoiitlirig conipoiir~~it normdisation.27) When we are workiiig irr t hc freqiicncy doinsin. (3. 1F for capncitancrs and 1 H for indiictttnces. i. W e begin with nrriplit ride iiorin ion arid niark tlir irforcm ci volt age wiOh U" and tlir reference current with 1.st __ . 1V.(.28) (3. also in SI nnits.on rjyc denotenormalised quantities by a tilde ( . the following consitltm are based on coiriplex frequencies.26) itrid (3. The simplest possibility for the ch of reieiceiice ~ o l ( r t~fc~reencc age.ioii i:. We obtain from . EIS m 7 d l as time constants in secoiitls wlieii dcsling with fast microelectmnic components. 162 for re ces. it follows that (3 .26) For tinw normalihation we relatr all timrs d to a rcferenw time tjirrierwionless x r o r i ~ i a ~ i s c ~ i time I.).ialfunri ion remains tliuncnsionless.%)). carried o u using (:<.t / t o = . all specified tiiiirs are expiewed as iniiltiplcs of a refercnce time. Since we want to represent compownts by irnpetlaiices. " t herefoi tfiswss irt detail . time normalisat ioii.29) Thc tmniitioii froni thc izrriplit iide arid time norridisation t o cornpoiitrit norrrialisat. Both . However. U'F' exprtth~ coniplex hrq~terlcy tlie likewise witJi the reference tinie to. WE cxprcss T I j 5 ) and I ( s )ili amplit udc ~ arid time normalised values with (3.Yto . Yuririalisdtion ran tliiis rried out so .unplitucle normsliiation jor volltlge arid cwrent aiitl tirnc irorniaIisat. We have to make sure that the argument in \. 1A and 12. Tim e normalisation causes a change in the time axis. currcmt r m tiinc is I .e Tiinc noinialisatition IVith time norrnnlisntion. viliic~s for capacities in IF are w r y unwir4tly.s i (3.From notl. ~ P oi lhr c c use onding SI units. Simply omitting these units gi3res a normaliscd representation which can be converted biick to a physical representation by ittitlint: ST unit s. Frorn enre values for the compo are immediately obtained.0 and obtain the (3.implitude normalisatioii. Firht. Both are dimensionless timeindependent vctriablrs. For riirront and voltiLgc3 in norrnalisctl Tmriables..29): .e.
32) frcqueutly occurring corrrponent 5 from Table 3.33) (3.34.) . Figriifinnctions 57 (3.3 '2.2 are: (3.
c = 0.01 oi Figure 3. From now on. 'Vhis represcntiition of the dinicnsioiilcss iictm ork is s2io.i?w in Figurc 3. we use Lolli varia~its t h c s i i : t l y s i ~ systems tvitl~oiitrleiiotiiig normalised and for of uniioi malisecl quailtit ies difreicritly. TJwally no sepiratc. millianiyercs arid seconds. The wniporii~ritwlucs WP simply givrn wi thont cliinerisioris. illodelliiig LTISysk ems in tlic Frrq~icncyDc7maili The system fiinctinri is then given by Ekie all quantities arc diniciisicmless a n d the result ol: iurtlier calciilation can he interpreted respectively as normalisatiou into voltl.8: Normdised RC' rircuit .. notatioi~for the aornialised yuantitirs (the tildr in oiix t a x ) is usrd.58 3 .8.
3. J(.3.( I) =  l'?. Exercises 59 Exercise 3.2 r 2t ( I sin 3 t ) A COSUl(]/ + H cos Zw.t Ire following signals in the complex fr equeiicy plane .$ sin(w.(t) (t) xercisc? 3.l t p) + = 4 13 siri( ~ )l t  ??? c 5cos(LVlt t 20") 4sin(I (It)  ??? 6  C O S ( 51 ) .1 Rwognisr t hti friqucncics of.
 L I .
ige at i .
'I'hc is region of convcrgcrm in this case corisists of a vertical stripe in the complex plaw. Ail ~~xaniplc given in Figurr 4. Laidarc Transform 1: Laplaee t r ~ ~ s f o r ~ i U Thf. tiiicl the path of iiitegrn. I (4. real nimilrer o determines the location of tile path of iiit.2) Thf. Lwplacc transform of Juntf/on~ ( t t) E IR 7s .1. It must lie wit Iiin the rcgiolr of ccxwergerice. but its cxact locaation and forni does not iiiflrtence the result. 1 r ( t )= P { X ( 5 ) }= L X(. .+"'tls .tiorr (dotted line) runs vertically from top to bo"mm.cgrittioii witliiii tlw cornplex plane.62 4.
1.r(t)} cii~i alw be itleiitilid by the syrribol o .3) 'The inteist" I~pl.6) t h m its valuc is zcro for t < 0.5) & ( t= ) 1 for t > 0 0 otherwise (4. expr arid (4. Use of thc step function will hr rxplairretl in the tollowing exsanqArs. . aiid it (*ailhe dealt with by the bila form as wit11 the iiriiliz 1 1.1.tc. F i g i i i ~ sh 4. (4.. sketched in Figure 4.Tliv fillcvl cir cle tleiiotca the frequencytloiirain quant it>). and this c m I)c infiriitc number of snprr iini>obe"d rxyoarrit 1 d as a Ric~rnanii sum: by tlici integral (4 2).c\ trtmsfororrn really d ( reprcsent R function of iirrtiorrs.2 furwtiomi iii th(1 timecloimin.aplace trarrsform.= C(.) .3. Unilateral arid Bilaterd Laplace Txaiisfc)rms I ~ 63 'I'he pairing of a tunc tiorr of tin1 ( t )and it3 Lapincc transfoiiir Xi.
4 les l a .t .
8) .I Iinjs) I (4.
Re{s} S~+3S+2 > 1 (’4. and a leftsidcrl signal for t < 0.4: Illustr.E(f)P t .2: #< z(f)= E ( f ) P .ttior~of the region of corrsrrgciice from Esanipl(~ 4. . as in Exaniple 4.1 0 ) YO’N we cotisidct a signal t h a t is not cqual to zero [or IX < t oc arid i\ formed froin a riglitsided sigrial fox / > 0 as irk Exdinplr 4 1 .Figure 4.2 X(S)= ~ 1 s+1 1 +  s i 2 = 2s A 3 .
11) Re($} zero loigurc 4 6. (4. . meam that t h r the regioii (Jf convcrgclic~ci i n thet strip lies between the two polcs: 1 X ( S )= ___s 2 + f 1 . 'Fhc region of com gelice is inaclt.2.3.i. arc^ in the splane in which both parts toiivf~rge.&(t)C2' . again ~ x p r ~ s bl lap lac^ L arisfonn i%\ its component par ts a ~ l d hrrl hit\ P t ~ I the silrrie forin. in the iritcwcttion of the two regions of (*oiiwrgcnccof the rxponcrrtial pails trctin Example 1.s: t 1 __ 2 s ?. ul) of thf.1arid Esarnplc 4..~1 X ( f ) = c(t)e' .I Im { s } M'r. Tlic t a c t that both regioiis.3 s2 + :is + 2 ' ' 2 < RtJ{*s)< 1. of convcrgencc of the individual tcrnis r)vc31ap. and so the same poles :%h in &hanq)le3.c. Illnstrstion of the legion of tonvergcrm 111 Exaniylc 4.
s First of ail wt' recall thc meariiiig of integrals VI ith i r i h i t e limilb of' iritegratioti arid (i.5). an integral 1 0 a r i d c)(i is tlelirred by the limit (4.12) .gra1 with finiic l i i n i t s 0 inid n :e m. LnlilrP a i iutc.
I x”(s.._____ ns I...... . .X(S(1) ..Asil litrr X(S0 t As) ..14) ..) (3.
aplacc Trannsform .70 4. 1.
wv recxaniinc bhaiiiples 4 4 and 4. In ICxaniylc 4..e Traiisform 1.13)) b~ p r r t together detl conipoiients. E x ( t ) has a finite duration and if G ( T ( L ) )coirverges for at least value of s. so irk t h i s case the integral (1. t h r pole of the rightsitled cuniporintt ( s = .sid s ~ ~ e ~ 91 ~ nsform of a bilateral sigrial c m (5ee (4.detl c o r a p o n t A i i t lie to the left of the sirigularitim in 1h r left . Rctanse 01 the properties UF 1ia. t. arid we clo rm ed t o hnti auy limits.vc discimed. Thc interscctiori is empty unlcss all of tlw singulnrities iu the right.thcu can iepreserit its I ~ p l trarisform by oiily tlici first integral ori the rightharid srtlc. properties we have jwt Iual rcgions of (‘0 germ’.the sum of a r ~ ~ h t .5. altliongli 5 (I) c:m coritam singularities its t i n integial o w r ~ ( t ) rlo ancl in that case’.. then the r e ~ i o of c o r i v ~ r g ~ i~ i e ~ entire ~~~~~~~~* l~ s the “c 0118) If x ( t )has finit r tluratiorr . i f .Region of Corrvergerice of the Lap1uc.1. In order i o fiilly imderstarirl this. esal .b i d e d component. Iri Exaniple 4. All of thv coiisicleratroiis coiiceriiiiig the legion of s a q .and left. there cariiiot be any siiigularities.5.2 ) 1ic. a i t d the corxtpletcl re>gionof wrivrrgcncr is t h i r iiitersertion ‘I’liis iriteriert ~ o i r s a i lies L o the left ctf the riglitniosf.s t o tirr left of (lie polc of tlie lrftsicled cortipoiicnt ( 5 = I) and thc region of coiivergelicv IS the strip Leiweri lhesc poleq.s of the rightsided cornpontwt) or to the riglit (singiiLLrities of Itlie leftsided coiqxmciit). should 1x2 clrar t h a t thc singularities of the Laplace tiaiisform c\ithcr lic t o the left or {‘tie rty+iri of convergerice (sitigii1aritic.13) dues not (~OTIVCT gc.s i ~ ~ e of c o n v ~ r ~ is a stre e~~ r i ~ ~ t .5. whcn r ( t )is oiily nonxrro for 4 < <: R. tlic polc of the rightsitled c . In the region (Jf convci griicc itself.1 ) lies to the right of the pole of Lire lcftdw1 (‘Ol?tpOrl(’Kl~ ( s = 2) and (lie iritersection is empty. ririgrrlarity ancl to (lie right ol’ {lie leftmost siiignlaritt.hc L forin for ~ ( ttherefore d ) .o r r i g m i w L ( A = .
e .
hiit with e.r(' intcic.icin.1 The function of time f ( t ) = e ( l ) and g ( t ) = c(t) fulfil1 bol.!lieu L { . m7e saw that fiiwtiom of t.xh diffcrenl region of convergrm'F". For the Laplace t. This meails that! th(i ucy$m of ~ r' i i anti U ..tetl irl knowing wlicthcr assigning R tuncl ion of t. If the following conditions for two functions of time are true: khen f ( t ) = g ( t ) rvcxywhcrr whrrc f ( t ) and y(t) izwe continuous.imc ( 0 a Laplace traiisforrn is iiriique or does this representation of tile function lose information that. so that no overlapping o w i m e(t) oa S 1 .tr.h conditions 1 and 2. the existence conditions for Laplace lmnsforurs derived from hc siifXicierrt but not I K y. largcr than the stiip With the ex. I!ic{cs}> 0 . but not necrss. the Laplace Lrmxifornis each have a uriiquc function of t iiur..y conditions for the convergence of integrals.Chapt 6 s 61.4. if tlrr region of ccmverge~i(~ coir is into cr)nsitlcrot. and likewise. cannot be recovered by the inverse Lap1ar:c transform.irrit> can be uniquely expressed bj7 their Laplace transforms. \\'c a.rmdorru. Iristead of cmmining 4 the dttails of Ihe proof. we will corisider two exaniples that clarify the conditions of the theoIerii. r ( t ) } x i h fur c' < Re{s) < D e As the estimates iiscd are sufiicient. lit. Exaniple 4. The lollowiug thcorcnt gives tkir ans~vcr.myles in Sectioii 4. F ( s ) = G(R). The proof of this theorrm can be found 1 .
arid tlrp tlieoi em confirms t lie observation ihilt all t hrer fnric.ion vahe at tlir tiriic. t = 1. They <ire 21 only thitinguislied Ly tlirir value at the d ontinuity at t = 1. Only the rnidclle f m c t ion c m be expressed by the step fulunc~ior~ (4.zrisloarm d o r b wtue lly lose I II fc )rniation: the infxmttt i on about iimct ion vahes at discontinitit ies. the tissigrirnerrt of a ftinct.msforais of thr three fmctictiis 9horjvs that condilions 3 and 1 are also fiilfillrd.1)) Calciilating tlw lay lac^ tr.1 f~ilfillconditions 1 arid 2. which is rriarltcd by o black dot. Thtt tlirorc’m says nothing about.73 4 Laplacc Transform . of hwitcliing t .7 shows thrcr furic*tionsthat . The h t e p fiiiir iion ~ ( t )I good csamplc a~ i t rcprescnts an idcd switch opercition. the tfiscontiiiuity at. be irnpossible. as I lie diflcrcnccs between cach two fimct ions for f < 1 arid t > I are YCYO arid the intcgral.0 is arbitrary. Adiiiitteclly tlioiigh.. that ib for t :c 1 ant\ f > 1. wliich woiild in lad. o w l tliern ako h a w the valiir zero. The previoiis exmiiplr sliows tliat thc rrprcsriit ation of a hlnctioli of timc by i t s Laplscc tr. tliscontinuous fimctioiis of tiriir dre often only itiralistd vcrsioiis of 1 apiclly cliangiiig coritiriuous sigiials. for which IS .  ~~  Example 4 8 Figurc 1. A .6) ( c ( t .tions arc equal evcrywlicw where they ale cwitinuoms.
c~lnr. Wlirii tl~alitig with cmuplicatecl signal? it i s an mlraiitagc.9(t)) exist.ncy plane.4.1) Lias berii tlw shortcsl way to find thc Laplace trarisforru. so in generd the conibined region o j con super. The region of corrvt‘rgence for the cwri1)incd funclion is tlir iritcrsectioii of I lir~ rcgtoiis of convergence for the individiial tmictioris. From the linearity of the intcy.1j . Properties of the Laplace Transfiorni . l i o ~ h . h all v. will learn the rtiost iriiportmt propertics We of the liitplncc~transform in tliis section.ioii. e v a h t i n g the i n t ~ g r d (4. difler1 t o pcrtornr ail o p e r d h n 011 it t imcdt twtiation) in the frecii~‘3rrc.g. and avoiding liavitig t o cvnluate the grnl clirrctly.tli~>s the co~npltlx i on fi. to be able to use tlic. Sirigiilarities irray br ic~iovccl I y the addition. as long ab both L{j’(t)} i d L{. J.20) i s iinmeditLte1y ohta ivicrl .rat. pxopcrties of the 1.7. set of the iritri on bctwecn the indivitliial regions of convergence: [ ROC‘{cif + h q ] T: ROCC(j { O r / y } l R i (42 1 ) T show lincwrity. (4. hottevcr. arid foranulate thpoicxis Ihi tlicrii ‘I’hc tlicoreiris will b r rieedctl in l a k r calculalioris usiiig I hc Laplace traiisforiii.ydoirIwin. y ( f ) is ixisci ted into tlw tlcxfinition of tht.75  so ee For the simple sigiials we have wrisitIeicd so far. Oftcn 1 signal (e.aplacc o t r i t n s h m (4. n j ( t ) + b .
.
4. i.1). s E ROC{s} (4.23) Thc region of convrrgeiice is scaled i n the sattie way. if thc3 riiiit of mcasiirmient of t iine is (*liangedor if sp(w1 of ii imgnei i c tapr (. s F ROC{rr(at)} i f S .26) . Ca e An cxpansioiz or conrprcssiori of the tirnr axis o(*curs.e.25) (4.for cxamplc.E I t O C { z ( f ) } . a (1.hmges. Properties of the 1.aplace ‘Transforrri 77 4.7.7. E’rorn tliv Laplace transform (he original ti signal X ( r ) = G { r ( t ) }. This relatioiiship is sliown through the sitbsLilulioori 1 = at’ into (1.
8: Sigrtai with arid without stcp at t = 0 Partial irrtcgration of the second integral expresrion yields (4.30) .78 4. Laplace Trarrshrm t Figiirc 4.
i t i s often given 111 the ai~d therefore hits tl x : fOK 121' I LJ{<+(t)}sX(5) =  r a (4. d 0 fi > 0 .27) <is c ~ ( / ) does riot have to bc diffrrciitiiLble: .6) . infliiorice the ViililP of the LA I diould also be notrtl that a t t = 0 the d u e u ( 0 ) is e q t ~ to ilie rightsickcl 1 l limit ot ~ ( tat) i 0.33) wlierc it should be noted tlrwt t h r fimction .firred at t =. as wt.h a v ~ alrcady see11 in 011 4. ('1 34) It expiebses t h r h t c t i o n ~ " ( t )which like (4.c(O+).33) We h a w now obtained the> cliflertmtiation throrciti ot the (1)iliXteral) Laplace t ransform for iigIitt. tiriilatcral Laplate trailstorm m. Properties of the Laplace Tranbfurnr '79 i t is the sitim' as the dcrivative of ~ ( twlicre .32) iepmimts ?lie dcrivtiitive of J ( t ) ~cLnshm . tlic ~ ~ n ~ l n ~Lwplnce f eral oiily for positivc t i m e 0 f < x .o hr t h e lighthand side > ) liiirit (4.is oitcii used.i ) u(0)= lint r(0 i.r(t) and its clerimtive is only of iniei for t : 0 mid tlic value of ~ ( 0 slio~ild\hp iinderstood I.6. (4.4.\trhes the integratiori tlicoiem for thc iinilatrrsl LC>p1~i((' tr msform (4. x ( t ) arid thr 2(0 t ). For iiniliatc>rai signal uiir differentiation theoierri (3. Tlw iiitrgratioii theorrrrr of tlic.34).idetlsigiials: C{ I " ( f ) ) = S.2 that tlc\iaiicc~sat isolakd Imints do not.7.I ( f ) r m be tlifcrciitiated (t # (1) ) a n d it is not dt.0. This vitliie is also kriown as x ( 0 i 1: . This h C J k in the defined regioii is iiot a problcm. of t ifre inilialh oiily tielirretl f o ~ > 0. ) is not clistingiiishabie from the bilateral Laplace transform.U(S) ~ ZjO+) ..33).
37) yields .t Usiiig the Laplace\ tr‘artsfc)vm anti partial iiitxyption (4. 4.29.30) for all of the suminccl tcrrns in (4.
t(t) 1 fi?.=I i.tc*etransforrrr of A(A)is firrally. .37). I ~ P Lapi.40) (4.32)).(t)C[f .. I I? . Propertics . 2 = 1.) = . 1 hat is.4 1) . (1.) from (4..the Lnplace Trn~~dorin of 81 Tlie expressiou in the s q i r m ~ brackets ol tlic hrst iiitcfsml corresponds t o the d(vAvciiniioiis ( w e (4.7. (1.39) \IJith the step heights S(b. twlirtliiig tlic 5tep: I1 &.4.
82 4. z ( r ) (t)c"'dt (4. with (4.45) dX(s) = ds ~ j.r).16) C{f r ( t ) }= _____ ) : dx(s cl s s E R0CI. Ltktplace Transform C { * r " " ( t ) = sG{.r"(t)} }  (J5T .42) and some rcasrangr merit. The l ~ ~ d a traiisforrn ol the triangltl sigiial is tliercfore oe (4. 1 (447) I .+ j2T)Z 2 (27 )Z ~  _ _ I 1 (LAY (2T)Z (4.13) As s " "( t) i s zcro a t d l poiiils whcrt\ it is definccl.
\Vith (4.1.48) (4.As a r t oxairylt we will calcillate thc Laplace transform of n:(f) = 1 ~ ~ ' ' :'.37) wc obtilin (4.43) .( f ) . I 'We will tie iising t h e trnnsforirr pair tlerivetl in Examplc 3.
_ _ _ _ _ ~ s (S ) te%( 1) tn C'%(t (sin (('OS j t ) c (t ) wot)c(t) c) r ( ! )= t'21 E ( i .7') d) . x ( t ) = sirih(2/)~ (t) . r ( f )= f C " l ( t ) t ~ ).
8 Uerivcx t h r trarisforni pairs irr Tal~lr 4. .7 .1) 1 = 1’ .23).T = s‘ .a. by sihstitutiiig (4.7.1 starting with t h r transform pair I X ( 6 )= It$>{ } > (I irsiiig the Ilieormnr from Scction 4.24)( t h e aiid frccpicncy scdirrg) by substituting t into (it. theorcm (3. = at’ xereise 4.r arid Exercise 4. Exercises 85 Exercise 4.4. i .22) b) t l i ~ modultxtion theorem ( i. s . .8.1).5 a) t h r shift. Prow eqitatioii (4.
This Page Intentionally Left Blank .
Thr integral (5. For V A < v < ~ € 3 . lex &(si = ( 2 7 e (5.raiisfoirn leittls to siinplt.o function t l i e o r ~ :Tlie ~I)P first step is to coinrriit tlie iiripoitarit r t w l t s from complex analysiq to i ~ i e ~ ~ r o i j r . thcx socalletl u m m e Loplaw fran. . As s may take any value in the compkx plarw. 7 1 r v e .1) (. S i d r an integral is writ1 PI^ (5. d s on the lrngth ot the palli 1%’.I + 3 Q 1 ( s ) of tlic cornplcx variable s.1 show5 such a paiaixietric ~ p i t h o intrgrntioii in thc iplane. For the reverse. carrying out the iiiverse Laplace t. likewise f o r E / = ~ ( 7 ) ) lakes thc value of the enclpoiiif [ I .2) means lhnt tlie vdue of &(s)ds is. TCJM n r tliis path t precisc>ly. For if = v 4 . eccuiiiitlt~tedfrom all of the infinitesirrial t~icmciitt.th x e given as a fiiriction of tlie real parameters I / .. Tlir real and imaginary parts of Q(s) axe denotrJd by Q . par. calculations that (‘an per for iirctl without having to continually refer hack t. all of the vahies on t h c path between llie start and cntl points niust be given.the vtiliies ot s that lie on the pti..I ) . s ( v ) follows the desired path of intc~gratiori 14.Figurc 5. i s ) arid Q l ( s ) . verified tlic Laplace transform.sjorni we gave the tyiiation (4. It will be shown that for these e.In Chapter 4 we started with t i n e furic1ion5 and tlirnugli valuation of the iutegl a1 cMitiition iii (1. s(v) takes thc coriipl~xvalue of U thca start poirit A. integration of such a fuiictioii it i riot siifficieiit to give s two limits of intrgratirm.urrr~f?ac. Tlris is called a. arid ham no sigiiificarit singularitits. ibe the.2) wlicre 1V is tlic path of iritcgratioii irr the complex plane.2) without giving a jusli In t h i s chapter we %ill look at the inverse trandorIn iii depth with the corrrplPx fiirirtion t henry LVP restrict ourselves to Laylace transforms that oidy have poles.
he lc~igtliyrxpressivri irr ( . tlic known rulc f o r real irkegrals is also i d i d for the path integral in the cornplcx planc . S(I1) Iltd by 1eal and (5.B  infinitesimalelement ds (5 For evaluation of the path integral ~ i o t only is Q ( s ) repi as iii (5.2) i b actually only B compwt forrti of t.1).3) iiito eqmtinri (5.2).2) caii IF vicwecl as a b i i m of i m n y real 54.3) = (T(Z1) I p J ( v ). value intcyqAs. ) 13s (5. Iriwrting (5.4) ‘I%e path intcLgra1 iri the cornplex plaiie (5.rttion = s(u) gives (after multiplying out ancl collectiiig terms) two real int cgrals with rcspcct t o the rcal (5. but also s ( v ) . arid usiug the sitbst it.
zliieof a circular inwgral coritainiiig R hingularity depends c m the roiitt) of the path of iriltgiatiori. we (an rctcr hack to the c a l d a t i o n of two .3. Nest we clarify whet her tlir v. howvrr. For the rdcnlatiori of thr difference betmecn thr two cwcular iritcgruls along 11.i two tliflercnt c*losctlpaths of integration 5$71 and 1.idar integrals that coiitain singrda1ities.t. region G. 3 whit11 si1ow. Circular Iritcgrah that Ericlose Siiigiihritirs 80 i o u s statcrnerit refers I i) paths of iiitcgratioii tliitl do not (~ritlosc~ singular isation of complex funttjoris. We consider Pignrr 5 .5. the singulttril ies nrr 1 relatitmships 101 circ.i within the analytic. Both riitlose a sirigulaiity that lirs withiri the white region in thr middle which docs not ktrloiig to G. aiid 1.I.
Tt> and IV:$ in Figiire 5.rr it.f ~ . integral d o ~ riot depend on tlic form of t h r path curitaiiiing s the siiigular iiirier regiorr.4) we (*ailsliow in tlrr ~liiiiit1 way that the circulnr intcgral aloi~g11 i s t l sanie i .3: Eyiial valuc integrals aroiiiid a siiigu1.’ ~ the stmi of.the circular iiitegrds around the iridiviclual siiigulaxilies (15’1. If tx closed path of iritegrat ion T I ’ coiitaiirs IrirdtipIe ringularirirs (see Figure 5. . thc va111~ a cir crrlitl iiitegrnl 011ly ctrperrds of on which singularibies are enclosed by the path of integration I V . t coniiwtetl analytic G . Coiiiplex Arlalysis .90 5. .tnd the Invcrse Laplace ‘l~atisform ~~ Figure 5.y This aigiirncnt can be used tor aIijT two riirg rritegrals around a siiigtilarity in G.4). T h e value of the.
(5.12) .9) analytic region 6 ‘ Y s (5.11 ) (5.
s)(5  s. Tlir particularly a..)].zte when Q ( s ) is a rational fractioii furlction.io~~ ~l a circular i u k g a l for a complcx fuxict ion with poles can be done iisirig the residue tlieoierri (5.1I ) hlthoiigl1 i t is not t i T I only Q ( s ) is klrc liriiit: XY.propri..e ?kansforrn With thr C’aachy intqyal wc can easily give Llir ii simplc pole ol Q ( s ) at sc..15) The ciiI<*i~la~.10) arid (5 Is). For (5.. \7dinc>s. a[ 5.9).l.. (5. cif tile resithies (5. tlw rt3siducs can Le found Iisiiig complcx integration. (5. I residuwii X. = lirri 5511 j&(. Complex Arrwlysix and the Inverse Ln1)lnc. To Cakiilatioii of the re e rtcrtl s ~ i i fiirthcr result? f r o m tire Cauchy i~ calculate multiple pole r .12) can also be foirntl tisirig the R. so far only bceii sho\vii for simple poles.92 5.
18).tl 1lie loc*xtionof the 77order pole sbL (2.19) a n d die correspondingly altered notatioii iir (5. wc obtain a f o i ~ ~for l a ~ iiiultipk poles that.17) Fyw) .l)timrs with rcy)c>ct o s gives t With the function (5.  (5.e. Deriving ( I I . being analytic) of the complex derimtives of P'js) cntailb regiilarity of F ( s ) itsclf. Then 1110 = p.rds aroiind sirlgl? arid multiple poles.?( S ) of (5.integral. iisiiig thc rcwidue thmcwr. st.22) ox\ showed that the Cauchy integr a1 s g i i ' t y simplified the emlinfcml uxtion of circular irltcy. 1 2TJ 1% ( w . i s e q i k d e n t to (5.13) (5.21) is again obtained without coniplcx iIitcLgrntion.mothcir simplification thst leads to a form ot the inverbe Lnplaw transform. Iiitcrchwrrgirig intrgratiort a n d drrivation with respect to s is permilteci as 1he iait.egral convergcs iiniforrnly. l n thi:. W e riow 147rik it irr tlie fbrm It can bc shrrcvii that regularity (i.19) (5.20) Thc residuunr . from (5. It caii be iisrtl riot only for polecs.S)Z dll* .ction we will discuss . but also for .
/ \ s \for sufficiently I large v.I3 ) j Coinplrx .iIuos of i s I (5. whidi ii. tlecrcasr tor large v d u e ~ / s / we i . . A i ~ ~ l y sand the Inverse Laplacc Traiisforrn is essential singulxities. We consiclcr adti makytic function F ( 3 ) within a region. of choow tire circular integral around s o iis in Fig. t1ia. which TUII:.t decays hy at lea~st. paxallel t o tlir irnaginaiy axis at i j clistancc crp > o ~arid~ l ~ c~ t path I1 . a n arc with radius X aiouiid (lie oiigiii ~ . we have l o rmke soiiie restrictioiis. WrJ. 5.24) with Llie pohitive r e d nurrilm A To use thir.6. rile malytic region contains all values of 5 where Hc~{s} > c ~ ~Thc path ol intcgxation consists of the path ~ ~ i ~ ~ . but first.
a n d a when F ( s ) is only kriowri along d the Caiichy intrgral. It is thcwforc.OK that w e haw ft c o n d i d d .gral i s impor twnt tor two reasons: L~ B T2ir sirripltfic*atioiiof the pittli of irit rgrat ion cwrisidcr ed in sec. E To derive the iiivcrsr Laplam Ir a n s h i m wc start with tlir (:auchy integral in the foirti of (5.jI (5. fwiction f.16) IS niovod t o location oI interest. T l i ~ F in polr i n (5. The condition (5.kiuii 5. It is likcwse siifficieiit to ltnovv F ( s) along a line parallel to I ~ ini.1) allows a simplified ( alciilation of the irtverx transhrrn (wit Inoiit complex integration) for systcms with single 01 multiple polcis. rompletdy sufficicnt t o know F ( s ) or1 tlir bortic. F ( 5 ) can tvem hc audyticallv conti*ruetl outside of closed path W . tlw Caudi> inityy. fimction has a 1 strong inner structurti.) i kiiowri along a closcrl path I T .~ ( s ) (vc~rywhcw 1he rnclowl region.1 of a n analytic rcbgion.2 l ( w h directly t o tlie formiila for the inverse Lnplace traiisfoim ion 5.28).16) can be rxspd to calculat. ti. 1x1 het. line Dccau\c it is ililrercLntiAAe.4.tgiriaiy axis (5 28) to h c able to cnlculate every valiie P to tlrr riglit of thr.5).. a 1 aiialytic. tlicl Cancliy iirtt. tlw reviem '\.4.~~~~~~~ oft If an analytic. "A?>will now us(' the result t o tlrrive the l(Jri1111lil for the invase Laplace iraristorrri arid give siiriple wtys to prrlorni it. (. aiicl F ( s ) i s arrnlyiic* i enclosed region (Fig.23) i b always fulfilled by a rational fiaation . In llris rt3sp .
inverse iiiii1atc:ral Laplace Irarisforrn we iise tht. w h r w we imitiiie s RS s‘ and so as s 0+7 x x.28). 1 F(s) 1 . Cauchy iiitcgral for thr a path of iiitegratioii parallel to thc iinngiriary ilxib (5.32) . Zri P(S’) .To d(qi.ic. (5.
in tlic . bv deiirritioii.cplitire (set.loc.34) .eral Liq)lace trnrisio f i r riglitsidcd sigiinls aircl the correspoiitlmg im p i se ials (see kJxxri1)le 4. 4. A s the rcgion of cnrrvcrgriicc for lcftsidtd time functiom lie. leCl o f a vrrtic a1 liii.32).rrc.5. E ig.4). ( h e positrx e orieiitation of coiiiplcx oritour iritegrali is ( oiint..29).ltwisC. so tliai me cn w re tli5tingiiishcd from each othw hv tlie tl gral arid 1Ire sigii o i t h lrftsiiltd function. Tlic invcrv hilat.5. Sjrice. l % i %lcittls to a change of i h c s i p to (5 29) for the Laplace transform o i leftsitletl rxponriitinl Rmc*tioii\ (5.2) Thci dc~iiwititmfor Icftsided ct out itb i r (5.tlw par allcl to the iniagiiiary axis 4ras t o iw tr. thr ii ciili~rintegral iiiiist d s o he contairied wilhin tliti it+i ItalY of the splane.29) to ('3.wcrwcl in the opposite direction to (<5.
.) . r“(‘Sj = .) . ’ 0.. (5. U’ithiii tlic region of convcrgcrrcc of L’(s). 5 4 . 1 .{. except tor the points i = U.i s analytic in the mtirt cmniplcx plaiic.I(s) : Et. i t i s possibl(>t o dioose it closed path of inkgraliiin with 1 5cld of icgnlaiity that crrcloscs thr polc s = .X) A 1 1cotitrast t h F ( s ) .1 1 (compare Fig.
39) as e” i i analytic in ihc critire complex plait‘. (5.c> ..)] Sb..iilatiori of the inverse Lc~plac.s)(s.cs.iL  .lirn [P‘(. C‘al(.
2.43) has a siirqlc pole at 5 1 .40) I *I' I I (5. fir) with = P { F ( s ) )= Hi _i R z (5. T h e imww transform t requirt.1 arid a double ~ m l at~5 2 .1 The Laplnce trnrisform F ( s )= (3 1 i l)(si ' RP{<5] ?> I (5.ctlat both thcsc points.11) (5.46) .41) The use of this torrnrila ih denionstrat ed iir ail ext~~nplc~.(5. ~ Example 5.6 that 1hc residues arc c~:Llcillat.
ial fraction expanbiori R ~ight&k~l .48) Wr can see Crom the region of ( oiivrrgciice that we arr dealing with frmctiori of time.Fmm thc two resiclues h e functiori of time ( w i now hr put together. but for coiiiplc~ poles. of ns1Jsirig iIir i ~ s i d i i c tlieorern tivckls e\alu. \r carried o u t p i ~ r ~ with real talculatioris.. I hc corresponding function of Iinie is likewise real. w'e will of r i aiid show the procethirr in an cuarnple. the iiivrrsr tvarisforrii leads to a viglit4cled Ciiiiction of t h r : .fit) = L 1{ F ( s ) } = [c'  eLf  fe"'I E(t) (5. As F ( s ) is iii latiorial fraction form and conwrgm in the right haltpl. cvcn wlieii corriplrx pnI pear (here irr c*oiiq)lex t orijugatv Cornl). An ex~miplc~ this is given in the next section.47) A rc4. Si artiiig wit11 tlic part.lne. FOI 1 ational flat tioii Lttple tiansfor rns with i e d niuiierator aiitl drnnminator coefficients.ition of complex residurs. We wait to find the irivrri(1 1 ~ ~ ) l ntrmsioiiri of rc (5.ttionship to thc partial fraction expaiisiori c m be ronstructcd l i c loo ~ for multiple poles. it rccpiirrb c*alcul. T h i s can he dolie using ail appropriak ly c ~ o ~ ~ ~ b i r r i t t i o partial lIttction c~xpansio~is tlie motliilation theorem. It would therelore br coimnicAiit if tlie i transiorrn cuuld I)(.ititioii o f coriiplcx ititcgrals.
102 5.58) .56) (5.49) into a form tEiat corresp t o siiie ~ r i t cminc function5 l of tinie.dil.B l)(. WP Ilricrefore waiit t o bring sccond term of t. C = 4.. so F ( s )(s + 2 ) 2 = A( ss I+2I) d + BI(S + 2 ) 4 . 42)s s 3.lie squ Bs + c 92 I s t 10 + 52 + 4 s + 40 6(s 1 2) '6 ( 5 + 2 ) L 36 . This prtrccdure also .' ( s ) ( s 8 1 + I)] = I.1' ( s r ay (5.39): A = Lirn [ I .. Substit uling s = 1 gives llie exact value ol A becmise the othcr two lorins becoiiic zero. (5.52) Piittirig this into (5.aiCA p i k .kio partial fraction espanuion (5. + (5.55) For simple poles the coefiiicienis c m be calcula~etl (5.ivorlrs for Nil.i K =.132 (5.53) F(s) = (5 + 3 A 1 . X'e (*anexpert that fur a secondorder denominator with rcal cvelliclrihs arid ccimplcx conjup. that tlic col rcbpoiiding f wirt ion of t imr i\ romp(mtl ol ternis. '2lliis can be done by coinpletiiig t.49) givcs the rcprrse~itatiorr (5.57) again.1 s 2 _I + L +.+. Cornplcx Anal) sis m t i the Inversc Laplace Transform we obt.
for cxainple.60) (5.62) ' = [e'  FLf f  6 .6.A~Nl(S+2) s +a1 y ' t 4 [I+ =Az d ('5 s 7 2)' I 'I . l m i in m i t i d the region of conveigeiicc: .1 gives tlic 5auic result that we obtained in Exainple 5.y E(t) .s) = (s + l ) ( s+ 2 ) 2 rn.1 with t h residue r a l ( ~ i h l i o nas long a we . Exrrcises 103 (cs a i d p u t in s _= 2: P(.5.f(t)= P ( F ( 8 ) ) '(5. + I d F ' ( s ) ( s + 2 ) 2 ] = [ d45 ds . . using TiLblc 4.61) S +I 1 s +2 I (s + 2)Z 1 (5. B 1 (5.64) xereise 5.
1.has 5 .L C .542 + 25 +.Exercise Fintl the irrvciw T.1 F(5) T R {s > .Sli r  A"  s . . 1)"s 4)' _ I _  + xereise 54 F(s) = .' ( F ( S ) } represerrt arid tlic resiilt as sinr i ~ i cosine tcriris.ztplace transform f ( t ) = Gl{ F ( s ) } o f 29 . C h i p a r e differriit nretlrods of' p r t i a l e } (5 4 . Find it4 inverse ~ a p i u c e tiansfoforiii f i t ) .5 +3 conipkx conjugate polcs m t l a rightsided rcgioii of convergriice.5. i a) 'Ilsc the complex partial fractioii cxpaiision F ' ( s ) = A s .
tlralt. the ystcin arialysis that. that ail 3 . $Le octtlinrd in Chaptrr 3 i s complete: .riise rigrirtls. to cover coarbiiiatioiis of L T b c solution ot initial value proltlerrrs will tliw ht. W i t h this.TIsystern from its eigenlunctions. h m e all it provitles an clcgant rl ription of the properties of i network tlirory.il coniponciits iit (he s y s t c m oirtpnt cornglcte outpiit signal i chievetl with thc invcisc Lap1 Io ii)u m I h p .2. Cloiribirririg t 17c ixrclividu. givcii initial valiies of tlie o u t p i 4 i g f l i t l IJS m r i he consictcrcd usiiig it. it ii the.. We :UP iiow c * o n c t ~ i i twith tlic drrivation of tkic system rcsponsr t o tdateial ~i signals m d with thir we (an determine tiro tirmsftir ( o r system) funct erid of the chapt(1i we will mtencl the rcsiilt.2: 1. itiiiidiird inetiiod for dcrivd In pnrticiilar. with in Chaptrr 7.Thc L a p l : ~ ~ transform introduced in Cliapler 3 does riot only srrw to i h r a c t. era s IVC have already encourrtcwd in C'hapter 3.
4) Frorn I his follows tlic ltmdaniental relationship bttwceri {lie inpiit uritl ontpiit signal5 of Tn’Isystems ( s w Figure 6.5) .1: Eigciifunctitxis of ail LT1system Inst earl of t LPRicrriaiin W ~ K IU e xiow write Lhc correspondiiig coiiiplex integral (6.2) (6.Figure 6.
In Chapter 3. Fiiidiiig the Systeni Fuuction 107 \Ye will tiow tlemonstrattl Irow to lirid it system fii:llnctioii.5 in the 7.5 we will wvrk with riormniisrd.6.1 t.3.01 0 0 . iwiiig two elrctricd ricltworks as marnples. wc will dviiionstratc the proceduxe with phpical uiiits. c = 0.thoris liere .2.in Esaitiplrs 6.1 ncxt chapter. WP will prrfor’orrn both of these iric.Lvaya of avoiding physical tfiinenxions.2.xampl~s to 7.o 6.3 we learnt two . a i d in E. ‘Thc in cornpoiir~ntshave iL1rpxiy been YIUInialibed S O t l i i b t t hcy lrxvr helpfid valiirs. The first iirtcvork w c ’ will coiisitler is the KCcircuit cirpic*tcd Figurc 6. equations.
8) .i  f (6.
7). A simpltl partinl fraction csxpausion itlid inwise tr.01 U.01 0 0 ! ! Ul(5)  1 s Figuir 6 7: Syhtertr fimctiori and system response of thc iiidcpriidrrit RCcirc i i i h fiorri F1gnre 6 6 Ilriving tlic circuit with a st rl) fuiictiori gives i t ransforrn of tlw output signal l similar t o (6.c = 0.~ns~orirtation of both ternis with Tablr 4.(t)= E(1) c = 0.1 y i d d ~ L .
Th& toration defines the ~ii111eof U j s ) (both rriagriilutle . Analysis of‘ ContinuousTinit.1 10 6.T . The zero at F = 0 arid tht. b .6).9: The magnitude of a system f i i n c t i o r i . LTISvstmis with the Lanlace ?’Iaxisform er ru clrst. Figure 6. t p 20 15 10 _. then thc corrrsponding system function will aiio bc simpler. pole at s = 10 arc easily rccogiiisr.3 tmd 6.9 shows i t s rriagnitudc on the romplex hequeiicy p l a i i c a . Only r a t i o r d fractioii ryslein fimctions will occur when itp number of eiisrgy stores. it the rircu neiits.6.7 / 5 0 5 IF 10 Re { s } Figure 6. Tire fiiiictions here cmi br tletiiic~lby giving all of their poles and zc’ios togethcr 01. I firnc*tion of the RCcircuit in (6. A plox of these points in tlw splancl is Itric)wn as a polezeio diagrnrn.ible.tnd phnsc) at all other points on the sylt~ne.It i s theiefore sulficiwt t o give these points i o tiefine N ( s ) to a conhtant factor. Emw so.riptioii of the ‘l’he systeiri fimct8ioriis if complctelv different h i d of c described fewer nrtworlts in Figures 6.
For polynoixiialr le~ with real coefficients.3.11 shows a sirnplcl rxarrtpl[J.2s +2 Figure 6 11: Polezero diagram with conjugated complcx polpq . they alwnvs occur iii complex conjugate pairs. I )  K .])(A . 5 ~ s2 I. Figiire (i..+ 1 . Polcs and % c w s of the System Function 111 H ( s )= S 3 k 10 Tm( s } x ~ pole 0 zero double E'igurc 6.6. I1n{s) X Double zero at s = 0 role> s = .10: Examples of' polezero diagram!: ratioiial fractioii s y h t ~ n i fiinctiorr call A t ) take r ~ i ~ i p \ialirclh.? li2 3 +. Iiowevm.1ic j \ I X I I ( 6 )= (.
niveigence of the systrrn fimctioii is 1101 normally of iiitcwst. As t l r e region5 of coiivergerice for riglitsidctl input signals arid cauhal systems always overlap. traiisforii\ of the fiirrc ti on of t i m c x . 'I'hr locatioii of ~ e 1 0 5 lias no infliimccx on the rcgion of coiicergenct'.usTinie LTTSysttw:.gion of convcrgeiice lits to the right of a vertical linc throiigh t h c iightltrnost siiigiilnrity. From the cliffererililtl equation a. Y ( S ) = 1O. ly vields the tiaiistcr function. 1Jse of t he difl'crcntiation throrem (4.26) xepl clvery tle~ivaliwof B fuiictioii of time by a prochct of <i powvc'r of s and the lac(.112 6 . iise of the tliffererit ixt ion t lieorruri ( U f i ) j k l d s thc algehrnic equation 2S"(<S)  (6. arid wilh xatmiml fraction systeiii funrt ion.SX(*S) 7 X ( S ) . Analjrsis of (:ontinuc.10) 3 s Y ( s ) + t . 1o thr riglit of I h c riglitixiost pole.12) .7.4.  (6. Thc gc~ice system fimd ions iii g m c r i d . if He( s } i s choscn to be large enough.11) (6. done with t. with the Lnplace Transforiii causal svstc'ms the rc. not just causal for cvtergc' later horn t2ie corwolutiou theorem (C'haytei 8.2). It is rvcxi cnrier to detrxrrririe the liaiisfer fiinrtion if an LTIsystrni is gil er1 as a tliffmmtinl eqwtioti (with constant coefficieiits). r n function was found ic~lizi iwly easily Ironi thc nc>tworkiii Scctioii 6.1/ .2 (\xists.wo exurriples. The tliflcreiitial eqwtion tliw becomes a n algc4x air c~qwitioii.c .the rcgion of coiivcrgeme of the inpiit signal X ( s ) and the furict ion I l ( s ) must overlap. WP will show how this i:. In order that the Laplwt~ rarisfornr of the oii~put t sigiial Y ( s ) in Figurti 6.2 re i i s e of cornplcx impcdtlnccs.37j + 571 = 1oi .
ii 20G 1ooy .wc' obtain the algebraic equation (52 t 209 t 100) Y(S)= 2 Xjs) .18) . io tlie tliKerel.15) + + ari xa x (s) = 3 s I.14) which using thc cliff crciitiatiort rlicorerii (I. (6.2 . (b.5 3 *5  5' 5 < Re{s} < 5 .eniial torrebponds c'q uat ior I .%).J7) Y (5 ) = X(S) s 1 20 S = ___ 4 5t10 3 s+5 1 s5 (6. (6.
3. " To show t.beROC 6 < K : i } < 5. The inverse tsramfr. we strparalt the ewes t < 0 arid t > 0. The third tcrm leads to B left.4: (6.3e.Y in Exmplc! 6.oiivergence (ROC) of Xjs) lks completely in . (6. y ( t j = c~~~ t <o (6. t?hefirst. For t < 0.The region o f c. t.Oy(t) = i ( t ) .siciedfunction o t i m e .19) fulfilh the differential equation..Iiat y(tj (6.12. z ( t )= :k?st.21) . e"ke(t).(t) . this. we begin with the differential eqnation of the KCkircuit that c a n be eithcr takcn directly from thc: net.wo tenw lie left of the ROC and lead to rightsided functions of time 1)ecauscof t'heir poles.tm c m take place as integration within {. so that n{.the region of conwxtern f'umtion for the causal RCcircuit Re{s} > 10.20) $(i)4 I.""E(t) $.19) The sigmls z ( t ) and y ( t ) are illtist.work in Figurc 6.ratcd in Figure 6. or alternatively from the transfer Ciinction B. To test. f y(t) z 4e"Ott.
6.25) .6. C'orribining Simple LTISystems 115 (6.
3): iq)i cwrrts the dcsirctl system liiiiction for the diffcreiitial eyrra(6.28) .iclerit ical olztpul signds Their prodiirt tiori (2.
14: A parallei swtern (6.3 I ) I . . C'ombiniiig SirnDlc LTISystems 117 output signals idciit ical Figure 6.G.G.
32) s Togethrr with the relatiotrsliip for wries coupling. 11it. however.33) c.igianis.tmplifiration factor V . thc transkr fuiictiort of ew. then the> a1)proximatioii 1 W ( s )= ___ C(S) (6.10 contains a f w t l (lie kiiid iii Figure 6.34) I f the amplification is very high..Every brarich of thc parallcl fmui in Figwe 2.. to systmis with onl> one input and m e \Vc now ivould like to extend these rides to cober by riis with rtniltiplt~ .11/.35) e s T l i c rulcs discussed in llie last seci ioir for scv ies< pnrcillel ant1 fccdback coupliiig of Liyst ems van alqo be combined a n d u s d to analyse coiiiplicated block circuit cli.11 path i s (6. The trimsfer friiiclion of the in(t g r a t ~ r s c integiation tlieorclm { 2 271. We ale still confind. The transfer fitrrtt ion o i thr fcrtlback system i s Ilicri (6. while the trmsfer fuwtion iir the F(wiback p ~ t h a is 1 F ( s )= Ci(12) = A.15.ollstallt: Example 6. 2 = 1. ( s ) = v . (6.9 I f a sphtem with transfer fiuit I ion k (s) is air ideal arnplifier with .
6.7. Comlsiriiiig LT1Systems with Zfultiple Inputs ant1 Outputs
119
We llii\cl a l i d y rnct sy ms with rnitltiplc~inpiits aiid outputs in C:hapof a systeni niitli A input s i 1 ters 2.3.2, 2.5 aiid 2.6. The trarisfei fun(. j m liich toinbiries tlir vectors a r i d l i oiitpiit sigiials is a hl x AI matrix )rs Ujs) of thc trairsfoiined ontput of the traiisiormetl iiiput signals arid the 1 ngn;tls (sec’ (6.5) ).
j ( s )=
H(sjxij_l
(6.36)
(6.39)
a 1t 1 1
t h erclor c
(6.40)
Iri contrast L O systems witli only one input arid or ii put, thcl itiatricec; R 2 (s) tarinot hc intcuchanged.
(3,
arid
iiiimbt*rof coluniiis. We can tlmi esprws the addit ion of the outpiit signals by addition of tlir systcrn fiiiictioris
Y ( 3 ) H1(s)X(s) I =
H2(S)%(S) =
ill1
[
(SJX(S).
(6.41)
This yields
(6.42)

outpiit sigrials
itlent ical
mist
Figure 6.18 shows a tcctlbnck sy m with ~riidtiple11iput5arid outputs. Here tw take special e m  ' with thr dimeiisioris ot tlicl nnatiiws. IYP begin vvith ncltlitictn t c;am(' 1~urnbc.r of rlrrnriit s I f input at tlie input. All vectors here i m s t hav vcctor Xis) has A r olcnwnts. tlien nintri s ) riirrst have AI colurnns. aid inatrix G(s) niust h a v e Af iows. At the output signal Y ( s ) fioitr I h r outpiit of F(s) is c.oiiibinrd with the input of G ( s ) , so i ) has K eleinerits, F(s)mist haw I<
6.8. Analysis of ShteSpwe Descriptions
~
121
arid
e
T l i c r i h we krii,vr roiisidcred ior the parellrl
systriris with iriultiplr input.;
itild
outpiits are cntirely sulficichiit for arlalysib o C
12%
6 . Analysis of C'oiitinuousTime LTISyst,ems with 11ie 1,aplate Transfom
many lirlsysterris. As RII example for rhcir systernatic use, IT e will calciilal e the transftJr fuirction of' a general slatespace repieseiitiitioii foi it systcrn with miiltiple s h o ~ in t h r hloclc diagram in Figure 6.19. I1 represents the n sl atrc~liiations ,234) described in C1iaptc.r 2.3.2. As betore, WP start with
a, wctor
x(t) witli AI illpiit sigiiids,
or y(t) with Ir' output higrials and
OT
z(t) with N si ate variables
pacr rrioilr 1 t 11(xi i Ilalrc diiriensi(xis
The niatrices ot the sta
c
b)
KxM,
I< x hl .
Pigtire 6.19: Block diagraru of
R
statespace representation
The coaiplctc trarisft
i;iinetl hy suitable maly
(s)f o ~ this statespacc reprc
icwt transf(>rfunctions a r i d rules lor conibining LTIsystmis stepbysl ep. Pii5t of all. we recognise a par idle1 nit in Figiirti 6.19. with two componcnt systcrns with the transfcr functions .5) and H2 (s) .
(6.47)
e transfer functions in itre direct prt.th nt. From (6.42) WP know that for the liar
(6.35)
The coiiiporwrit transfer functioii L ( s )inr~st deterniiiied fiorii the lower path. be 11 cotisists of t hie stems i r i series (6.40)
(6. I9)
6.9. Excrciscs
with
~ + ( . S ) =
123
c.
(6.50)
‘rhc component. tiansfer finrictiori I122 ( 5 ) i s ubtairicd by analysing the fewihack sjhteni t’ctilt consists of t,he loops 6 ( f ) ;uid ~ ( t ) . the notatioti with
1 F(s) =  I.
s
G~s) A
(6.51)
we obtairi with (6.16).
T is a unit) matrix, sizc N x N . Now all c.oinpoircnttxsnsfcr functiuns h a w been f o u i d . WP obtain irorri (6.48) ;tnd ( 6 491, the cornplctc tiansfer finction is
(6.53)
o x will1 (6.47). (6.50) arid (6.52)
(s)
=G
[SE A]’ 
(6.54)
BY applying the I iiles tor conibinirig lirlsystems, we have foi.~rttl funclnmenthe tal connrction bet wecn the mitlriccs of tlrc st a t e  s p x r model, and tlw tuinsfcr fiirictiori o C x system with rnult iplr inpints and outputs.
Exercise 6.1
Dcteriiiinc the transfer fnnction H (s) of the follo~ing series resoriimt circuit iising coxnplrx irripcclnnccs.
L
R
R=l
Exercise 6.2
Ihww thc fnnc(ioxi u ( f ) = ~ ( t ) f arid t h c tmgeiit at the i)oiiit,~ ( 0 )The tangent e . caii i i h e ( l to help coiistrix*t thcl graph (if cxporrciiiia1 functions. if its point of intcrsectivri with the zaxis i s ltriowri 1) mine [lie point of iiiterbectioit .
Detervriiiie tlw transfer functions
N(s)
aiid &aw thc polt17c~rodingrains of tlic
svstcms.
xercise 6.5
1)ettrminc. tlic tranbfer fiirrrtiorr arid 1he tliffercittial cqiiatiori of thc systein v i t h tlre CoLlowing polezcro tliagraiii and H(O)  1
126
7.Solving liritial Condition Prohlems will1 the 1,aplace 'liarisform
.lsl Classical
The starting point tor the classical soliitiou to tht. probleni cicscribecl tlbove consi ot calculating t lic sohition of tlic lioniogenous piohlern a i d of a particular soluti of tlw nonhomogcrioi~~ problern. A gcrirral solution is lorined faorri tlrr b n i m o f tlic tao, in wliic*li p cvioiisly open pararnetrrs are cMncd, so that tlir soliit ion the z lulfills the initial contlitions. T l r ~ follownig rxriniplc shows this classical soli11ion tor i firblorder sy l
The liomogeiious solution yfL mist fulfil1 t 1 . 1 ~diffcvritial cquation r?)!, t ) (t) ( 2 y f r ( f = 0 and in this case it i s )
I
\Tit11 the c.omplcx aiiiplitudc I' riot yet deteriniiic~d.hisert iiig iitt,o (7.2) yields
7.1. FirstOrder Systems
with
127
(7.18)
7.1. F'irstr Order syst>cms
I29
Tlir initial wluc i s tlw d u c of the oulpul sigclul itt (ime I  0 For sigrirtls wit11 a step at t = O. this is the v i h c of the righthand limit y(O+), regni(iI(~~s of wlrct1tt.r or nott thc plus sign is written.
2.2. First orclcr yystrrxi in dircc t forrti I1 m d t h c timc constxrt ir = KL'. Figure 7. tisiiig the dilkrentiat ion theorem. the differcnlial equation IjLz t.ardnce Trsrisforrrr This shows tlial i I i e intc~rrral part is deterrriiiirtl by tirt\ initial state of thc inicgrator iii Figuic 7.130 7.( 1 2 = i l l ir (7. we obtairr . and froin that.23) followb immediiitely. Solving liiitial Coriditioii Problrirrs with the 1.
27) . FirstOrder Systems 131 (7.7.1.
32) ..Y(s) = Ycx+(. ( s ) i (7.sj y.
35) (7. tlie inverie trariseii WP rel~resrn~ a i i t partiitl hwctioii.‘I lit. .c trarisiorni.39) . l’ol wliic*liwe t tely For the f i r i t k r m l<:xt(s)= H ( s l S ( s ) .ind t h t calrulation ot it I bc simplified by f i r d OC nll r t y i t LIrld S( j with p 2 (7. oiitpiit signal is given by thc inverse Lap1nc.
.13) it ~ioufollo)ws that the ouipiit sigrinl foi t 3.iilI ot tlre classical solutiorr (7. of thrrtl parts 1s 1 rxannplt.r (1 e"/?. wr ran that ('1.44). bor a siriusoiclal input signal. wc expwbs this rwult.36) (7. i t s ronsidcr further divide rip llie external term of the by.44) To c. in thc saii~e forin as thc rri.44) consists .are collected togctlicr. U i5 dt)= exterrral p a r t (7. Solving Initial Conclit ion Prol.orirhrdc.lcnrs cz ilh the L a p l x e Transforrri M'itli (7.43) T h e terms with .13).1 34 7. giving arid thc rcsuil 1s then pnt into (7. The h a \ rvsidt corrcspcinds to (7.
s the transitlnt part and the p 0 l c 3 ~ of llle exuit.aplace txmsform alltws the respoiise ol air EVlsystem to unilateral biglid\ to bc> dctrrnii~led just as elegantly as with bilateral sigiials. i. As . Vse of tlic (31 H’erentiation lieo or ern for rightsidrd sigirals t rriits a input signal at I = O a prrdetermined initial \iillitl of thc out1 i he resporrsr to a rightsidctl sigiial can l ) split into ~ exterrial tcrm r 1 a1 1 itrterriill a d . the irittrml part.i s pcriornird b y partial fraction t’sp into transient anti cm*i pole of tlir system ( 3 .s ) .44). 7.17.\lri~acly xricntionctl d m e .r.~o) detciinine tht? exritat ioii part. the tlecay part i s idmt ical t o t 11 oiid term in (7. tMerrr1iiic.III .iitiori ( s = i~.. pivaary T h e results of our intensi\ c 1111estigatioii of firstorder initial condition pioblerns can 1x1 summarisd by t l w iolloxving poiitts: ’The 1.
txid (7. (7.4t tlie niiclrllr srnioiliirig iiode (compare (2.54) yicAlt1s .53) .136 7 Solving Initial Condition Problems with llie linpliic.53) substituting in (7.15) for N = 2) IEiflr\rc\rrtiiLting (7.e 1kaiistr)rin .55) .
Figure. idth<)ogti the 1 [wilting expressions are iiot always as sinrplc as in (7. 7.58) Rrplacirrg initial values by i i r i t iiil s t a t r s is also possihlr with ot Irer struetm‘es.58).3: Secoiidorder system in direct form 111 (7. s .
z ( 0 ) .tor tlic first clc. Used on the state vector z ( l ) it. i s C ( k ( t ) }= S Z ( 8 ) . 3 3 ) .tlie t i m e bchaviour of ihe state wsiwe R ~ C lookiiig foi the output aides can b p drterrriiired il it is of intcmst.The result is ~ wt' r/ .( ct kiaw any iiitcrnal st ructulc.rix:ttivr.62) e o tlic matrix (51A) must s t a v on tlrt. ~ e Tkansforrn ~ VVc start with tlic iirtlrr N system with oiie input and oiic o u t p ~ t t i.61) (7. y ( t ) . w c ?ors to (Rj . left 1)rc. p l .~ ( 0 )= A Y(s) = @Z(S) DX(s) 4 (7..138 7.tj3) into (7.ausctlie iriatrix producl k 13y i r i w r w Liiplace transfoiriiiAtioii (7. c ~ i sod siibstitutJr (7.63). H s i g i d . aiid so also any eoefticicnts for mslriws To me the Latpliice timsform o11 t l i i s vystem of N firstordcr differciitial eyiiwt ioiis we only nccd t hc differentiation thcwcrii . Solving Initid Condition Problems with the I ~ .59) 'I'he Laplace iratisform here rcf~rs the individual component\ of the st att. however.6%). ('7. givcri by that the statespace drscril)t ion ( 2 .
.63) describes in a gcricial form 11 lkrorri tllr input to thv output of a stiltclsyacc Tliis prohlern has :heady t> be irittt. T h c rrspoiirc of the with initial cmnditioiis to a i i input sigiiill t ( / ) c . i t is ji .4.5) = C(SI L___ I I 'I'he equation (7 64) has thc initial state&c*oriibirirtl HS ii how t o cal(w1dtc 1 he tlarlsfel %triicturc.64) is illustrated in Figure 7.rpreted RS a vector of t.. H ( s ) = C(sIA)' (...7 .ransicr r i b tlir t r i i ~ ~ s1A.SpOi1st. from the Inatiiccs structure w (7. X ( i . atioii (7.T h e N initial states z(U) m i s t snch t h t the coiiiplete solution y(t) = g.t h t rt.58). H ( S ) oy.21) and (7...1( t ) ..t(f) of n svslciri j oy.wioliOur from thc f~r tlividual intcgrators to thc output of Ihc ct Iiolth ) is the vectoi of the tiansfer tunrtions from the s 1x1 uf tlw wliolc s y s t t m . Nqiiatioii (7. (t)+$/lllt(t) fulfills tbr iiiitid ( . HighcrOrder Systems 131. 3 . that arise a i t h r intcgritlor outpritc..
Figure 7.5: Block diagram with transforined system rnatrites .
3.73) .3) is: (7.70) (7. ( 5 ) c m be intcrpreted in Figu a vector of tIaii4br functions from klie 11 It of the irrtttagrator to tho output and (7.tiS) caii 1 ) directly n ritt eii down from this. forni 1 1 in Exmiple 7. 'l'liib ih illso triic.65) and tld rrrriiiriiig it froni R differential equation.6. as slmtvn in Figiues 7.2. statespace tlrsc.73) (7.71) (7.7. hi with ally number ui iriput s aiitl out puts. The A firsh ordrr svsteni is rcprcsmtcd ixi diic stdtedescriptioii ~'2111 tlprir vcl directly from Figure 7.riptiori of a stwridort1t. HigherOrder Systems 141 . ~ ify the cwmectioii btwecn clctermining the sFst pnr fiurction from a rtiirc as in (7.5 and 7..69) (7.r syst erit in direct form JI I (see li'igiire 7.2: be (7.72) The.
33) at I = 0: y(0) = CZ(0) + .66) gcncratc.LtC z(O) is not avdldilc. chosen.64) to (7.77) Tlie qitatioris (7. s t a t e .oulput rclationdiip l o r systcrns of any t1egrc. Prom the outlint equatiorr(2. we firid through clifTeieiitiation and use of the state equstions (2.t h r initial st.nti>grs of tlic.34) in the limeclomitin.With (7.s p n w dexriptiori. wc also iisc t h r ?tdv. a coiiciv foriiiid I of thc input.e in a gcmeral s t dircct form 1 or 1 1 ktatespace desription i:. ho 1 1 rf matrix no~atioii.I (0) . It is c*oniirct ioii Ix+werr t lie iiritial In o r t i m to do so.
~ h results aic' e siimrnaiisect in matrix lorni ( 7. ...ivt.Afier thc N ....80)   0 . 1 __ ___ __  AA I I    .s of y ( t ) at t = 0 have b e e ~ calrrrlatd.1 deriwt. .
87) .(7.
7.93) (7.. HigherOrder SJ bieirii 145 (7..3.92) { 7.94) .. (s) = ~ ( S ) Z ( O= ) (7.9j 1 I.
7.3. HigherQrclcr Syslriris
147
~ x a ~ 7.9 ~ ~ e n
\Vr arc' looking for tlic output signal y(i) of a given input 9 1 g r d .I ( t ) :
A
firstorder systeln for 1,
b
0, for
0 + 0.I g = s,
The svskni firriction is
g ( 0 ) = 8,
x ( ! )  U9.1 siril ,
t
3
0
.
(7.99)
N(s)=
___
.'j
+ 0.1
1
R.e{s) > 0.1
(7.100)
by splitting X ( s ) H ( s )into paItial fractioiis:
Y w ( s )= X ( s ) N ( s )=
(93
10.1
t r,js +0.1)
__ __
A s f0.l
c
Bs
52
7
+C +1 .
(7.102)
The cwfieintts arc o h t a i n ~ d either as in Exarnple 7 5, or bv equating cnefIicwnl i anti solution ol a linear 3 x 3 eqimtho svstem. The result ii: il = 10, 13  10, C = 1. Iriversr trausformatioii yields the external part:
T h e general form of the internal part is obtairrrtl from the dcnomiiiatoi of t tic systcni funrtioa:
1;nt ( s ) =
s
(2
+ 0.1
0
pint(t) = ae".Lf ,
t >0.
(7.1114)
Thc cc~riipletesoltit iorr is thc sum 01 both parts:
If fillfills the ( ~ ~ ~ € e r t'cpationiand the initid conditions for CI = 8. T h r Fig~~ri~ ~l ures 7.7, 7.8 and 7.9 show Iht' inteinal part, the exlmrlal part and thr cornplrtcx solut,iorr .
P'igu~e 7: Internal p n r l of thr systerri response ixoiii Example 7.0 7
I
I
I
0
10
20
30
40
50
t
60
70
80
90
100
Figure 7.8. Extcrrinl part of tlr
30
r  7
I

r
I
1
7
1 
0
10
20
30
40
50 t
60
70
80
90
100
Figure 7 9 Corrrplete solution of Kxairiplr 7.9 .:
7.5. Exercises
tion 7.1 .I).
149
with thc Laplace transform completel) in the frequriicydomain (Sectioris 7.1.4. 7.2, 7.3.1, 7.3.2).
3 . system itrialysis with the Lnplacc traiisforrn in the freyIier~c.yclorrr;i~i arid calciilatiori of the iriterrd pail in tEir tirrirtlomain (Sccl ion 7.3 3 ) .
The classical solution is cei tairily the most camplicatecl nic.thoc1 in terriis of cakuIittion cdfoit. as it works cxclubively in the timeclorusin. It i\ difficult to guess a par6ictrlar soliitioii for higherorder svstenis. If the Laplace transforiii is nsed t o articular solulioii, thr classiral solution tiirns into the$ third proccd~~ie. ly in tlw f i rquoncy11 :i~inlysis with t he Liiplacrl !ransforrrr coriipl the inost s i i i t n M e procwlurc i t the internal cttue of the systc’m ih known, for cxaunple, an clecshic*alor prcseiitntioii. Then tbr iii\ t i l k 01 ! , h ~ iiiitial vdLi(’S. ierndl part c m be d&wziiined cithci tr The transfcx 1unc.tion G (s) aird i m t i i c ~ ~ s :ire obtained directly tiorrr t h e tespmxb representation. Cnlritlatiort of tlic. Intcmal p a r t in the timedomain i s witable if t1w iiitc~iial structure is unknown, and orilj the difI‘tJreiitia1 cqiiation arid the jnitial coiiditiorrs have beer1 givrn. ‘Tlir II stiry calculation steps are simple t o carry oiit. if the miinei ical values of the tliffvrcrrtridrquatiorr coc4kicirts have bcm provided.
ises
a) the hornogenous solution ytl. ( f )
1)) a par ticitlar solution g,(t)
(a)
tlir c*nmpletesolution y ( f ) .
IndicatP tlic intprnnl arid c>xl crnal part s of the ~1iitio1i
E x e r ~ ~ '9.3 se
Soli e the init i d cwnditioii p ~ o b l e ~ ~ i
.fj(f) i 3y(t) = . x ( t ) ,
>c 1 r ( t ) 10 cos(&), c > 0
f
d(l+) = &I transform. Considci tlw initial conclititxis usiiig IJie rliffei.pn{ iittion tlieorem IOL tlw nriilitteral I ~ p I a c c transform (4.34). its ~ h o w n Sectioxi 7.1.2. in
Del eririirie
a) N(s)
h) Y j s )
c) y(l) for f ;0 
Exercise 7.4
Derive equation (7.16) froiii (7.15) itsiiig tlw differtiitiation theorem of the Laplace tri ansform (4.34).
Exercise 7,5
Thr followiiig fii stordcr hystcm i s givm a's a block diagram:
z(t) =
E(t)
'
(1  t)
+ E(f
1) (t  1)
40) Ueteriniiic
20
a the Ini(ia1values y(O) aiid #(U+). )
7..5. Exr~i5c.s
I _ _ ~~
151
b) thc output s i g u d g(t). Exercise 7.6
I’he following sccondorder sysirm is giver1 i n dircct for111 111:
X ( lf )
2.1 (0) ?l(0)
= .it) =
2c2
 & ( f  2) &(t
a ( 0 ) = 2b
a) Determine y ( t ) with (7.58).
b) C m (7.58) nlho be used, if the block diagram is also given in dirwt form II?
This Page Intentionally Left Blank
Example 8.1
The f'miction of tinre
X(.S)

L { J ( t ) ) ~ =
s
+ (1
a
.
(8.2)
154
8.Convolution and Impulse ltesyoiise
1
1
Cltwly ~ h lap lac^ tliansform of thc signal r ( t ) = n e O t , c ( l ) and thr s y ~ t r m c fririrtion ot the RCc'irciti( liave thr sarnr form This lrads us t o asttriw that relntioirships rxist lierc that W P have not yet dibcovered. In orcler to slickl soirir light (111 thcsc rclationhips, we v\ ill first ask sonip q1wst ion?.
ten1 Irmction N ( s ) be arsigncd
H(b)
ii
fiinction of timc hjt),such t h a t
= L{h(f)}?
2. 15 it possible to rxritr a sy tlir tiiiic function h j t j as signal look like?
n with a particular iripiit signal tha i output signal? What would Ihis sp
3. Can ii system mith system fuwtioii N ( 5 ) also bc uiiiquclj icleiitifird b y the c*orrPsl)oridiiig function of t i i r i c ii(t) if it exists'! What docs thc func*tionA \ f ) say about a system?
4. 1s tlierc. an equimlcnt, iclniioiiship to Y ( s )= H ( s ) S ( . i l )in thc timcdoiiiniir,
b e h e c n y ( l ) , x(1) R I l d h@)?
Tf'e will i.mYwt'r t liesr (111 ions
111
Iht~ following
Figure 8. RCcircidt wit 11 wctaugulau input signal I:
Figtrrc 8.2: Time bclinvi0111 of the function h ( l ) from eqirdtiori (8.4)
((I
= 7)
t
.2.
To flxplain tlir imraning of the fiiiiction I ? ( $ ) , we rvaliMr the response ot thc RCcircuit, to a rcctatrguliu irnpiilsc s(f)shown in Figure 8.3. Thc function of' tiinc and i t s Lc2place transform tile
0
othcrwisc
a r i d froiri the iiiwrse T~placc trwiisfonn, we obtain the otitput signal g(1) it,self
(8.7)
se .5 2 t Figure 8.8.5 1 1.s Belore 'i. that yroduws tlic response y(f) = of' h(t).m.espoiise of t l i c KCcircuit to different rectangle impu1xt.6:R. we liecc1 soirie iiisight irilo tlir naturc~ tlir input signal .n 0.5: Various rectmgitlirr impulses with unit . I with yes.3.Tit aiiswei I lip second question in Section 8. T h e Dclta Iinpulsr 157 Vigilre 8. c ( f ) .
is of little iisc a h it d o ~ not explain how b j r ) caii be niathes iiialictilly coinkhetl with other fiuictions.F ( 0 ) To 1 (8. m t ) for t # 0 and grows beyoilcl all limits at I = 0. F’irit we .~l>aiidoii the delta impulse in favour 01 the rectangle impiilse ~ ( t ) ( 8 . It is iicccssary to use tlie idea of di. 5 j .slr~b?6f~ons i~isteatl[17.___I 158 8.he delta impiilst. (8. Chirwlution and Intpulse Rcsgont. httcmp(iilig t o tlefinc h ( t ) by a riile that assigiis a .11) . The delta impulsc~d(t) is no( a iiinction in the usual spiise. WP mill refrain from a Ixiat. Instead of clescrihing the tklt a itnpulse by >iri assignment (for instarice (8. rat her by tlreir cffrcts on furictions. T h e rcsult is called the delta impulse b j t ) liiii ‘fil0 r ( t )= 6(/) . ant1 other dixtiibiitioris is (hat t licy ot described by their undetermineablc properties. Dirac tmpiilsc.e froin (8. Dirac dclfcc junc Lion.) .10))swe see whether the value of the integral tlescritm thc dfect of tlie delta impulse 011 the iunction f i t ) . and ztnd r m p u l w .lluc SjC) L e w i y time point t leads to a o functioii that tnkw the value . The cz~bual formnlatiori (8.4 0 With ( X . f ( t ) J ( t dt ) =?I* li f ( t )d t ‘ 1 F(21.v.10) although not wrong. and instead we will illustrate some iniportant p r opwt that will be iisefiil wheir dealing with sigilals and s ealt with using the aiialogj oi iiiliriitely t lriii ective erty The pririciple for iise of (.9) It i s also known as tlit. 5 ) . evaluate the from inlegral arid carry out the limit 7. A mathcmat icall~7 precise definition of the delta impulse as a reliable fuiiction for which evcry value of an independeiit variable i s assigned a firrctioii value is not posdde.liematicwlly thorough formulation of the dclta iriipulse arid related distributioiis. 191.5).
Note that f ( t ) utiiist be t writiniitms at C . a 1 ~ 1 ~ 7 q t i f The clclta. fctllow.14) This result is clrar if the delta impulse is iniagiricxl to be ilie limit vahw of o unit thc delta impulse ari arcn or bet7pr. The limit T o the differential quotient of F ( t ) at t = 0 ant1 yields w 0 leads to f ( t ) z ( t )tlt = liin I 0 0 x.7descrilrw this sititation.13) tztic property of the tlelta.iacl the clclta inzpulsr renzuves all values f ( i ) l o ~ # 0. It nieans that 6(t) rlisapprari for i J 0 and grows I)eyoiiti all liiriiti at 1 = O (compare with (8. inipulw i s riot restricted to sel iiig a \rii..0 it’ rlrlta. F ( l b ). it can be i w d to wlcct any poiut.9) we ohtain I I (8.1:i) U .ltic at t = 0. t Figure 5.I ’ ( 0 ) ~ G __  (8. arrow poiiitiiig iqwaxdh a t t = 0. we lmow imnicdiately from t h selective propcrty (8.13) that ~ (8. impdsc is rcprrseiited by ail Figrirc 8. impulse.12) f =o W i t h (8. iri fact.7: Sclectiori t lic fiinvtioii value f ( 0 ) If f i t ) = 1 b’t. It i i i ~ m s This r c k i ionship describes tlic s intrgral of tlrr prodtict of a fiinctiion .JU)). 1 (8.xlrere F ( t ) is an :uititlrrivi~tivPof j ( 1 ) (b”(t)= f ( $ ) ) ..15) . E. In thr samc way : in equations (8 11) t o (8. arid selects tlie value f ( 0 ) .
be t k d t with w l i m it is within an iiikgritl. Its proprities (‘an hc cxprcwxd in terms of the eikcts of d e c t i v c propcrty c m othcr functions of timr. Convolution and Impr~llsc Response This gerreinl form of thc selective property is ~ h o w n Figure 8. h(t) is thercfore called the zrnpwlse iesponsf~ the RCcircuit. we will srimmarise the taiiir4 so far. is airiilwisonw. is slouily rcl{ H vcdiiig itself to be i pc)w~rful in iclentifyiiig l tool its properties wr iieed ~iiore infonnetion aFout the delta. They allow iis t o answer the secorid question in Sc !A<> C ~ U S C 1tCcircuit depicted in Figure 8.8: Srlectioii of thr function value j’(t0) B~lore considering turt1ic. .1t o p o d i m ~ to the tlit t~ oiitpiit. iriipul The following deriwi ions of caltilation riiles for I lie ( k l t a inipiilsc drpcntl on two pIiiiciples. Wc will inwstigntr the proper ties of ddta irnpril Ariotlm possiblity would be using rectangle fiiric4 i 70 + 0. The iripiit signal t h a t does this is t k r ~ tlclta iinprilse 6 ( t ) . ‘ 1 % ~ tlclta impulse c m orilj.160 8 . e Calculations with the delta irnpulsc miirt be consistent with t’rw diilatioii rules for ordinary firnctioris. howcvc~.( b ) )that was prcvioudy anonynoiis. its WP did when deriving t h c selectiw property(see (8. w will stick with thc cle $0 v property. ‘. to The funclion h(t) = .8 in f Figurc 8.r properties of the t k l t n impiilsc. ~liis limit.
19) it IS iniiiictliafely yieldccl when CI = 1.9. T h e Dclt a Irrryulsr 161 rty of d ~ ln iinpul t is thcir behaviour M lrrri atltlccl ant1 o tlw f o i i r r d ( f ) + b h ( t ) . The sralrd tlrlt a inipnlw fi om (8. that the delta iinpulsv is eicn. Fiom (8.17) (8.19) Delta impulses with wigbting f cxs arc dtipicbtcd ab a vertical arrow sliotviiig thc c*orrrsporrdiiigwright.d(t)+ bb(t) = (a+ b ) S ( t ) .3.19) has t lit. I (8.T h e p f igatecl using t lie srled ivr prop x h 21 1 Cr. b o (8.20) . rrpresciitnt iotr shown iui Figurc 8.8.
Despite this. The investigation of tlic clelta impulse with the selective prop car1 bt.t of two dclta impulses 6 ( i ) . To tjlist ingiiish this from differeritiating ordinary functions. the term derzliutton is used.pretwiited by the effcc t. for exicniplc.fion.Corrvoliition and Impirlse Hebponsc Figure 8. (8. howevcr. of a delta impnlse 6 ( L ) . Diffwcntiat ion of S ( i ) by forrrririg tlir diffcrential quo~ieiilis of C O I I I S ~ not pc)ssilr)lc. s j t ) is not peiinitted.lzrinot he differcwtiated. Tlir dcriwtivr of the delta irnpulse has the same iiotatiori as the synilml fot a derivative with IPspPct to tiinc. be used on distiibrrtions. espandctl to other properties. we will again use the sclcctivr prijperty.21) This s i ateiiient is only true just like selective property (8. Under the iiittgral. thc drrivativc of d(t) ran bc transformcd using partial integration of thc corresponding continuous fiinc. a comparable operatioii can. iri fact. a s tiw tlclta impulse c. wri 11 h ( t ) .13) for tuwtions f ( t ) that are contiiiuoiu a t t = 0. x r . In particnlar. the piot1itc. aitd (11 ivc p r o p i t y of the mdefinctl fiiricrion 8(t)c m be re. forming a dcrivativc. ‘&I explain what bhauld bc iinde1stood by this.9: Scaled delta inipnlbe ~ ~ l t i p l ~ c a twith a ~ Q i i t i r i u oFimetion ion ~~s Froui the selcrtiw property.162 8.
13) and (8.11).10 (bclow).using the s r of Tht~ tlrrivation of h(t) is revcaled its dfwt on tlie contiiiimus faiiglc functions with a limit 1 o illri as tliey are iiat cliff crciitiable.24) . f i t ) . The Delta Iiiipiilse 163 1 d(t)f(t) clisappcarr for t 3 I lN only O i negative dwivative of f ( t ) L . We m n iiitrrprcxt the derivative h(t) as c? distrib tlir va~iie tiir drrivative . it h The corrtymrrdiiig limit for R bellshaprd impnlse fiinction cZ( L ) with c h r acteristic width T is illiistiatrd iii P’igurc 8. b r e ( w l d . however.23)). lim d(t) T+ 0 I bt (8.3. lime in(r functioii with a. limit.10 (top).lie dcrivativc h(1) of 6(t) show~i F i g u ~ e in 8.J ( o ) from D timction . In contrast to thc funct>iorr. LVe only ch impulse as tuiotlrw inipu fiinction I W C R I ~ ~ F ~ i i p icularly simple to irilegrate (8. x Ee at t = 0 remains. exact description of distrihutioiis is only p fiinctioiis (see (8.8. thc double impulse upvartis and dovlmvartls it i s d s o c d c d a douIt shoidd also be emphasised here that the graphical rcprrscntniori on thc riglibhand \itit> of Fig 8.llir dcrivative d(t) of the bellslritpd functiori can be kxxmL for T + 0 theu !teach to t.
(8. I 6 T)dT = ( I’ h’jc)t(t .z ) d z . (8.25) .27) .
1 yuot icnt (not 1 the function itself !) i i i c tqiial at tho point of tliscontinuit. I i a m .12 bottom lcit).liitli diffcrriiticztion o f ilie continuous part aiitl drriv.tl. rlie discontiimitj wii li s t c y liriglit U nppears in thc ckvivatiori as a &>hairiipulsr with wcight ( I (Figuic 8.12 buttcnri right).(8. function (Figiirc. Both t. nhicli bns 110 deLivativc. Tlw diffeieirce at tlic tliscoiitinuit> 'C.eirxis l x put together to givc the drrimtioii of the discontinuous sigii. in thc iioririal sensti.12 shows a iigrial ~ ( t ) left). klie step fiinctiorr.31 ) Figure 8. 8 1 % top riglrt).v. thitt carinot be clif€eirrrli a t e d brcausc (top it is tliscvntiiiiious at oiir point.ition of. be rtywscnted cis lie sum of a fiinctioii that can bii tliffrreiitinted arid U. Forming A cleri\ative in the iiorrnal way i s riot possiblc. 1 t8hr rightsided m d left. iiowevrr.1 I : D(41ci impulse and step function __ Figure 8. the desivativr of the con1 inuous p x t a n d it delta impulse with wciglit (I is obtairid (Figiirc 8.sidrd limits of tlic diffeientia. the signal r ( t )can.
d the l d p of tlishibut tonsi.l~m~nt yields the velociLy 111 t ~ I P form of: a itep fuiictioii.14 top).12: Arri~lysingsigrials wit 11 clisroiitiniiitics (8. and again we nc1c.pli. 13: A j)itck sliding on ice If the purk iu imiforinly set 111 motion iii. Derivation of thc di.f signal with discontinuity I t  Figurc 8.32) Figure 8. e couid ating tlic rlisplaccvrent twiw. Repeating the ck~ri\iitioir then yields forcc in the form of . Tltr h r . tlic proItortionally with tiizie (Figuie 8. f = 0% Iiowvrr.. t = 0 bj itn icc horkry stick. if y ( f ) tould be clifrerentinted. this ih riot the rase.
Thr tlelt the hockey st i t k . system rzioctel of LTIsv t he freclucnc3. CkJnvolut ion a tirltil impulse. 167 iiripiilse ticrr i s tlie idealized strilw of thr. MF coiiltl show only rising the dofinitiorrs of liiicaiity arid tiiireinvariance that fimctions of the fox111 P q t arc cigenfurrctions of UI'Isvskms Iy to the svstrrn fiinrtiori as il. .dornhin.t time iii descdjing the trariskr properties of LTIsystenis.4. piick with observed displacement velocity 177 11" y ~ t w i~ t ~ l ~ In Chaptcr 3 we W ~ I P intcrcsted for tlir firc.8.
TW now n (let i w an equivalr~itrelat ioilship tro (8.36) . of all U" describc the LTIsysteiir with its reaction t o I/([) = S{.2 as an iritrotlixttioii.*i thc t irricchnain.1) in tlic tiin~donrain. 'I'hcn U'('' USE tlie scl hip111 sipill i+h it fiiri (8. ~ v 2 clefinitioix of linearky and timPinvariancc T h e rrsull i b tlrcn gencrnl arid does riot only apply to t h network considered in ion 8.35) (8. Corivolutiou i d Irnixilso Resoonsc r we idcntitiecl the function o f 1ilnc1 h ( t ) T.. LVP will strart with uii LTIsyslern with it kiioxvn iinpuisc rcsponw h ( t ) and we warn L sllou' o )oiisc f r ( t )i s suficiciit to dcscribc the system's retxtioii to any input sign ionship brtwwn the inputs rial and output signal looks in a.tioii t o a ctclta i r i i p u i s r is h j t ) .Cl{ K ( s ) ) as anotkier system description arid iriterprrJt(d it as the sy I respoiise to a delta irnpulse.To do this \e only iieed to iise thc itnpulse respoiise h ( t ) of an LTliystcm. ~ 1 selective pi opcrty of tli" dclta iiripulbe.(t)} 1 input signal (8.3s) In particulc+r.168 8.S{6 ( t ) ) .the rrac..
169 (8. frorri the timeirrvarianct aiicl (8.34).3.38) lornaiii i s givcii by t he convoliition of the input signal with the.8. Convolution Finally. Ir’( 5 .
In (8.41) It is riot oirly tiiie wlwn h ( t ) is a11 irnpilse rrsponse and biit also gciierally for all pairs of functions ot tirrir wlr The rcgion of ronvcrgt~iceof the coiivolutiorr piotluct tcdiswtion between the iegioiis of coiivergcii~e RUC{ rcgionr oi ( ' o i i ~ r r g e ~ i c ~ overlap for the Lsplac~.s) = L{rh(l)}* (8.i (8..~ 5 c ROC(lz * J ) 2 RO('{. howevei.r(t)} does irot exist.ti miist product to exist. poles are cancelled oixt by wlmr C { h ( t )* .r} can. Expressioii (8.c}fl ROC{h} _ _.40) I h(t) ? X(t) .401 U ( Iixd set oiit the coiitlitio T' variable s can oiilv take talurs f o r wliirh both X ( s ) aiid H ( s) exist Tlie region ol corixergcencr ROC{lz * . it rnc"aris the convoliition prodiict (8. for rxainplc.42j . actiixllj he I x g w than the intcrscctiorr of RC>C(x) and ROC{h) if. 0 IZ(s).Figuie 8. ran br derivc~l from the corinectiori between the system fuiict ion niid impulse r~spoiise H(. .16: A systcai in the t imrtlorrraia (8. Thc JtOC{h} of the systeni fiiwtiori U ( s )that we h a w bo far not paid niricli erition to.Y(s) .38) is in riot a r i improper iiitcgral tliet only corivcrgrs iintler ccI tain conditions.38) t l o ~ s rxist itself.
8.4, Goiivoliit ion
171
Tlie iioriiicll rules governing the regioii of the Laplace trazisforro apply in this ( AYC, and as thev are all give11 in Chapter 1.5.3 we will not repeat tlierii 1ier~. Aii iittei ('sthg consqiieiice can bc obtained from the coiivolutiovi tlieorein ( 8 . U ) , i l w delta iiripidst: is input as ~ j t ) .
(8.43)
As h ( t )M 6 ( t ) = izjt), however,
(8.11)
TIiis caii be irnrnrtliatr4y confr iried with
C{cT(f)} =
1.,
x
elective propcrty
fit = I ,
&(f)
P
ROC: =B:
I
(8.45)
the entire ccrinplex plane, as the delta, impulse) The rrAgiori of cmriwrgciwc eiiclo is a signal of firiit,e ditration. Just as thc number 1 is the unity clement ol in1dtiplication. is the unity eleriieiit of mnvohttiori. Figitrc 8.17 illiistratcs this.
Figitrc #.I7 l,T[sy
tn
with a delta irnpulsc a input function
With the convohitioii theorem ~e tan d s o easily confirm that e z l is thc~ eigeizfunction 0.t LTIsystems. 111C3aptc.r 3 we derived (,his p ~ o p ~ r from the cuncepts ty of hiearitty a n d timeinvariaiice. Convohttion for s(i) esr gives =
J x
J  x>
(8.46) Tlie response t o an exponciitial function is likewise an exponential luiirtion. mnltiplied by t h f hys~ l'urictioii N (5). 'I'Cic~ corivoliitioii integral clcarly converges cmi when the coniplex fieqiiciiry of the the inpiit 5 i g r d r ( l ) lies in the rtygoii of couivergcnce of the byst erii lunction.
L 72 
8. Coiivolut iori ;uid Irrrpulse Rrsponsc
t
Figurc 8.18: The c oxrwlution integral
Thc iiitcgraird F t ( f  r)r(z is obtiiincd by rrialtiplication with s(z). ' L ' ~ Prt2c.t) angular form of .r( T ) m ~ a i i b that the integral can urtly have ~ioiizcrovalues for
8.4 Convolution
I
1.73
C'&SPS to
O<
T
5? . ;
Hcrc tlrcrc are t h r w
be distinguished (Figure 8.18 bottollr).
For 0 < 1 < 7;) !r(l  c) and s(r ) partially overlap integral rlepcnds on t. ' r l ~ rcsult is
SO
L h c k upper limit of thr
0
0
(8.47)
For. To 5 t, h(t  z) and .it) overlap completely arid the integration must be> carrictl oiit l)(%wcrn0 and 7;). obtain: We
(8.49)
ils we ol)t ained in
. ~ ~ d uirig tthe cwivoliil ion int cgral in the t hedomain is significantly more a
the frequencydomain calculation in (8.7). F$ccau.;e oi I he
c.omplicntcdcvcii €or the simple case c.orrritlered hcrc.
8.
~ i ~ t ~ ~ ~ a ~ ~ r s
First of all we would likci to know alxtut tlrr plopertips of a system tliit l l ~ s the. step fimriiori E ( t ) as its irnpulsr ~ C ~ ~ P O I I S C . the convolution of cjt) ancl z(t), fioiii wc find that with (8.39) tlic oiitpnt signal is
yjf) = E ( f )
* .r(f)=
x.
7
E(l 
zz 71 tlr . )(
(8.50)
(8.5i) wci c m leavc oul 1he p fuwtion iinder the intcygal in (8.X)) avid in over the values of T for ~ b h r c h ( f  E ) = I , that is, between i
I t y ( t ) = E ( t ) r ( t )=
h
1 1
( z) d z . .
eh
(8.52)
A systein with the response ~ ( f A)S inipulsc~responsc tlicretole leads to an integxcition ol the inyA, signal, or in slrort: it i s ail ii1tegr:ctor. This rewlt w;ts also found iii Section S.3.4.3, wlierc V.P said thilt iirtegrating it clelt n iiupulse yirldrd a si,ep fitiic.lion. Figui 8.19 shows a b1oc.k diagram of irn integrator arid also i t s resporihe to a delLa im~mlse the input. The coriwlutioir at
1) 1
Figure 8 19: An integiator rt?sporrtls to a delta irripulse w i t h a s t c y Cuiictiori.
8.4.4.
rent iator
1 1Sectioir 8.3.4.4 we got to know i ( t )as (,he dcrivativc of the delta irnpiilse with 1 respect t o time. K e woiilcl t lrtm+orc guess tliat i ( t ) w o u t ~ tlic impulse response ~ e for a diffrrrntiator. We can confirm tI. this i.; tlie case iF we f ir it1 the cwivolutioil of an input signal .!.it) with the sugge tl iriipulsc response:
r
? I ( ! ) = &(t) J ( t )= +
J
m
8(L  .jJ(
z)dz
(8.55)
hi (8.55) we ustd t h c selective propmty of i ( t ) (8.23) as wdl as the L d I h a t WP a krruw 6 ( f ) is odd (8.26). Convolution with the inipulse iesporisr ti(t) yields tlir derivative of fnrictiori L*it 1 resp 1 'Yhc system h r i t * tioii tor a rlif3ereritiaf,or is
0 0
II(s)= L{R(f)} =
J &jr)r"tdt =
i x
5
~
9
rC
(8.56)
W e ('an i i u ~ irvwrite the difl(~ieritiationthcorem of tlic LiLl>lacp trailstorm (4.26) dircctly as a special ( c of the convoltitioii theorem (8.41):
.I
(8.57)
htercstirigly rrio~glr car1 also iisc (8.57) oil sigiials with tliscoi1timiitie.s or wc on dislril)iitions, il we intcrprrt i ( f )the drrivation of .r(I). which niav c'ontairi its steps. delta irnpnlsc~sand clerivativrs nf delta iiripinlses. 1 car] tlim avoid liaving % to deal with the initial valiie x(0 I) srpaiatcly in the tliffrrriitiatiuir thcorcm for rightsided signals (2.34) o i tlir htep in Ihc tliffkicntiatiori theoirrrr (4.41). This js illiistrattd in the following rxample.
i 76
8. Convolution and Iurpulsc Nesponsc
The result,
i s then inimrdiatdy obt,aincd, i3nd it corrcspoirds to (4.15).
.4.4.3
Delay Circuits
tlw input signal at A delay circuit is a fixed t i u w clclay t o , rt from the delay, e signal rcrriains unc circuit to a d d t a inipdse h i t ) nrnst l x lh(' hysterrl v~'sp0"sc' of a d impulse Jchyed by I'o (see Figure 8.20)
h ( / )= S(L  L(])
,
il
ddta
(8.58)
8.1. Cixixolution
177
$
1
=
d i t  U).
(8.61)
An iiripiilsc train is shown in Figurc 8.23 (bottom left,). Convoliition of a signal .it) with ail irripiilw train 1e;rcls to a periodic repetitioii of x(L) with period 2' (see Figure 8.23).
T
2T
Figure 8.23: Coiivolutioir!wii 11 tin inipulsc irain
8.4. Coiivolution
179
rrietbods werc at first (starting U it11 st rani engirres) clevelopet1 v1)arately for each of the tlifFereiit areas (mechanical, electrical. cheinical engineering. etc.) but it was soon recogriised ihat the control procc.cliires for very different piohlenis all followcd the same p inciples. Ways of dexribiug systems indeperiderit of' their x imp1errreritatit)riwere then investigated, t h r goal being to repxesent only the sigiiificant rclatiorisliips bel,ween caiist' arid effect. It was customary in coiztrol engineering to use the p response t o clmw nis iirsteacl of the impulse response. This i s partly because the theory of ns wa? developed later, bnt also brcmise tlic step response is easier l o sponse. Tlid is untlerstailtfitble as not PVPI y niensiirc tlian the impul tecbnival or biological s can (lea1with a, powerful spike (i~s approxirriation an for a dcltn irnpulse) ab as tlie piick i l Figurc 8 13, l Thci rnost important elements of coritrol engirieeririg are depicted i i r Figtire 8.24. The gra1)liical symbols each sliow the funclanierital behaviour of the irripulse reThe formular for the impiilse rcspoiises are shown next i,o thein. The Icircuit is an integrator with a factor iii froiit of it. It responds to a step with a sleadily clinibing ramp. Its impulse resporise is a scaled step fimction. The Icircuit defines systems that can store wc7rgy, for example, capnvit 01's. The Pcircuit is a multiplier lhat i s defiried by i t s factoi . Its impulse resporise i s a scaled impulsci. A YIcircuit coiisists of a Pcircuit and an Icircuit connectet1 in parallel. The step and impiilse responses are cach the suiii of the corresponding functions for tlic. Pcircuit arid Icircuit . A PTj circuit repr cscnts a dilfeiential equation o f the forin
$ ( t )t y(t)
1
T
= a(f) .
(6.62)
An example is the RCiietwork from Figure 8.1 (tlo a constant factor). T h e resporise expoiicntially approaches a value th;lt depends on a tiriir constan 'The impulse r esponsr is a, dccayirig exponerit ial liinction. , 127'1circuit represents t hc differential equation 4
g ( f ) + ,y(L) T
1
=i(i).
(8.63)
p rcsponse shown in
An txxample is the NCnetwork from Figure 6.3 w i t h thr Figiut? 6.5.
.4.
ln Section 6.6 we investigated thc corribiiiation of siniplc LTIsystems ihat were cwinectcd in parallel OK serie.;. We saw that the comnhinaticxis cif 1,TIsysLerns led to niore complex LTIsystems wiiose system fuiictions could be f i r t ec2iily by oii9 cwnbiniiig the component's iiidivitlual system funclioris. For impulse responses of
f'circuit
h(1) = c l d ( t )
PIcirtirit
Figiirc 8.24: l , i r i c w , Imeinvariant control circuits
(8.64)
For the pa~allt.1 coupling, with (6.29),
1
H(S)
= N l ( S ) i Hz(s)
0
h(t)= h l ( t ) 4/ z z ( t )
1
(8.65)
became of the linearity of thc Laplace tianc;form. A sirriilarly sirriplr r.el&ioiiship for. the impiilse response of a f(:edbaclc b y S t m i , t h a t would correspond to the
8.4. Corivolutiori
181
identical
oiitput signals
Figure 8.25: Syaterns in series
i dcntica 1
oiit,put signals
(8.66)
i~titl1he complete syst,erri step response fox the parallcl coiipling is
For t hc s c v k coupling, howcvcr, the step fuirctions must be ronibinrtl in accordaiirti wii h
= $(tj
* S l ( t ) * .'3(1)
= S,(2)
* S 2 ( t ) = s,(1) * S2ir)
(8.67)
and their the rrsult can be dilTerentiater1. Alterriatively. biit it is cuinbersorne for funclions that are rlrfiiied by mull iple cases. a5 tEic convolution integial takes a tliffereiit form tor each case. To tiemonstratc the process.ignaIs from Figure 8.3.182 8.27 are defined by the func‘tioris 1 o5t<a 0 otherwisr 2 arid y(t) = o<t<a 0 otherwise (8. Tire coniplete step respoiisr is ohtaincd by tfiffcrenlisting the component step I or the complete prodixt of tonvohitioii ( N .4.67) can easily be expanded to cover N cascaded systems. called convolut/on hy insperf7on. because it bririgs an intuitive Luldeistandirig of the c oiivolutiori operation wliich is esseiitial fo1 practical work. _ I I ~ x ~ r ~ i 8. O ~ V O ~ ~ t ion by In Section 8.1) h i e s . both sigiids rmst bc fulictkms of .27 and fiist evaluate thc convolittion integral as in Section 8. however.68) ‘To calculate i.4.6 of‘ coiivdution by inspe~ctioii The two signals from Figure 8. Con~olut~ion Impiilsc Respoase arid One of the two step responses miist also be differentiatetl hcfort conivolut ion takes placc.6l e p \ t *2 4 1 Figure 8.3 WY biicfly dealt with the ctdcul&ion of tlw couvohition integral. Equation (8. This iwthod always works.he conk olut ion integral. ‘ r h ~ reader should try to acquire this skill. we will tonsidn tlre two rectanglP c. rzs it is not necessary to cwduate tlir conivolution integral for tlrcse sections. and w e b must bc calculated indivithially In this section we will desvxibe a 5implc mctliod that is well suited to signals with corisl ant scvtions.27: Examplc 8. With sortie practicP it i s possible to find the convolution product with this method. the convolution of the step response can be carried ont.
2 5 t < 4 I t 2 2 0  otlierwise 2t 4 . Consequently. it likely that there is a simpler way of pc~rforrning this convolntiori.70) 2 5 t r 4 4<t The corivolut ion integral takes t he vdues t J2di.he simplicity of the result. Convolution 7 183 and additionally.z) i s zero for certain sections of t: t<0 . I t is therefore sufficient t o orily calciilate the valne of the convolution integral at certain poiiits. but because wv hat1 to consider various differe s it is uririecessarily cumbersome. Example 8. To explain we will repeat the last example with the graphical method just described.r ) cha1ige. Tlie method of calriilatioii we have shown hcrr i s correct. the product f ( r)g(t .9 .9(t x % 0 4= i 0 { 2 O<z<t 0 otherwise 2 (t2)I:z52 0 otherwise O<t<2 (8. t2ic convolution integral gives either tlre vahc~ &er0 (when fir)g(f .> . The iiiterinetiiatc values of t h c corivolutioii integral call be obtaiiietl by linking the points by straight lines.r) can also orily have a (*onstark valiic for some sections. The first property that is not. one must be iiiverted aiid shifted in Lime: The protliwt f ( 7 )g(t .f( .z) = 0) 01’ a liiiearly growing or demying fimction.27.4. 1.is that because the signals f ( t ) and g ( ( ) are piecewise cmstant.71) This defines 1he triarigiilar signal sliown in Figure 8. 0 05t<2 Y(t) = J’ Sic) Y(f  4 dz = .2t 0 O<t<2 2 <t< 4 : ot Iicrwise (8.s.icealr)le. 2di. where the value o f f ( 7 )g ( t .ooking at l.8..
.
~ Pignrr 8. Coiivolution 185 ~ 0 1 2 4 5 ~ ~ ~8 . I . The characteristic poiii t=I overlapping starts.hr following results. . 2. which is.4. I Jbing OUT prescribed procr.29 shows the convolutioii of a rcctangle signal with a nonrt signal.dure leads to t. lioowcvrr const iInt in sections.8. ?‘hr signal g ( t ) is t Iic si and so it will br rcvi~rscdand sliiftcd.
CI(ir)d? = 12 = 6 .).186 8.)g(ir)dz  12 . the rectangle no longer covers aiiy part of thc higher part.29: Convolution by iiispection (Example 8.f(5z)g(z)dt=3 T . a t =5 t =G t the lower part of f (z) is completely covered by the rectangle. the rectangle stops being coinplctt~ly superimposed o. Convolution arid Iiiiaulse ResBunse 1 3 4 5 6 7 Figure 8.. By m d t iplyiiig the signal with constant sections at the characteristic points..)g(z)dr & cv t=6 J’.= 0 =3 %. Deterinirriirg tlir areas gives thr: following va Iucs fol the convolution iiitcgral: i x t=l f j m f(z)g(z)dt.ver the liigher part of f ( r ) and starts being part idly superiniposed over the lower part of . constant products are obtained.3= 9 N J’ f ( 4 . 3 . t= I t T 5 6 s” f ( 2 s” j ( 3 0 0 . =7 all overlapping with f (z) stops.9) t =3 f =4 t 1 1 rectangle is cornplctely superirnpt ~ the higher part of f (z).f( z). aiid oiily partially overlaps the lower part.
Applicatioiis .9). l\iIxking thc\ points (I. The convolritioii of these signals willi ckgrccs 1%arid d s to a convolution product with order K L 1.ercsting uses o f signal 1)roccssing are hasrd on convolution relationships.31.onvolutiorr. rnatched filters arid dec.12). 8.29 it becomes clear that the length of the convolution product is equal to the s l i m of thc lcnglits of the two signals. as in Figure 8. The coiiroliition ot two triangular signals ( K = L = 1) give5 n thirdorder convolution product (I< L 1 = 3). The points at the start and end of the corivolutiori product iire also the siiiiis of the iridividual start and e i d points. (5.5. by usiiig point 2 of our procedure on the start arid end points of overlap for general signals.29. Comparing Figures 8. by polynomials. .8. (6.3). Figure 8.30: Length of the convolution product of' two finite functions T h e prescribed procedim for signals with constant sections can be used on all signals that can be described i on!. Here we will cxaniine two exariiples. It can casily be shown that this is generally true for the corivdiition of all t h e bounded fiirictions (not just those with ronsOarit sections).0) and linking them with straight limbs forms the result depicted iii Figure 8. (7.28 and 8.12).27. + + + + Many irrt.x 187 e t7 f f(Sz)y(z)cJr x = 0 4. That ~iieanslhat corivctlution of a t r i a n g i h signal (I{ = 1) with a ngular sigrtal ( L '= 0) would have a corivolnt ion product for rnetl from secondorder parabolic sections ( K +L+ 1 = 2 ) .())% (3. (4.
response derived from the signal forin ) U ( 1): h ( t )= rrt.72) wlierr the f m r t i o i r m ( t ) is known hut thc lime lo by ~ l r i c h nr(t) is drifted i5 iiot ltnown. i~ . so t.practical t a&. the aim i s to iecognise knoRii signal for 1115.d y(/). we could trarrsiriit wri izconsttc~ 51gi1al? i t ( t ) .tllcv~ m i z t ~ k e dfilter. 0 The integral owr thc always positive d u e m2(r .73) The rniiius sign in m( t) cancels out) the reversal of the signal iii convolution.he o i t t p t signal is (S. ognised as the peak in tlie sign.75) ou . objcrt recognition 111 inuriy sinciple: In tlre sirnpleist rase wc \\rilnt to recogni5e when a sigiial corrrponent 7rr(t) occiirs.to) must be at Itthan y ( t ) cxccpt a t f . (8. the ieerivrtl signal s(2) is filtcwd with a n LTIsy a11 irripulsc. Suilable choice of thc sigrial m ( l ) can nialtc thc wlur of y(t0) iiiix*h greater than all other vahes of ~ ( t ) and thcn the unkiiowii tiiiic .31 shows the (8 73) arid a pc>ak igerneiit of a filtcr with ilnpnlsr respo this kind is c. For exarnple. tlrteimiiiih tlie ~inlrnc~wri time taken for it to retiirn.a i d with r ( l ) . which irieaiii identifying wliat point in t irirc thry emir and distinguishing titein irorri otlirr signal forrrrs.(tj. Exainplcs of thcse tasks are sp 11 recognition. To achieve Iliis. That niearrs we :ale dealiirg with B sigiiA of t h t forni r ( f ) m(t =  to). j8. Figiire 6.t o .
Applications peak detection 189 m(Q Figure 8.or recognises if and ~vlteiiits i signal form has uccmrecl. c. which g i w s the stiliice betwecn itself arid an obj all t. distance sensor Figure 8.10 As ail exaniple of a possible application for ii matchtd filter.39. btlt instcatl that an nlrcatly filtcrctl sigiial is rcturncd to its origiriial form. work out its speed arid c w i i i t the number. The problem is to decide wliat kind of vehicle is passing. Somet lilies thcx problcm docs not r q u i r e lhat a signal is ckia~igcdby a hlcer.inteiidetl to couiit ancl roughly measure the n speed of wbiclcs.ion with m . we will corisitler the arraiigcment shown i Figure 8.8. of cach kiiitl of vehicle. thc maxiinuin dcter. .31: Finding a signal pattrrn / n ( t )using coiivolul.r(t} hoin the in Figme 8.5. i s the couriting frinrtion with t.tl s niid lorries iiioviiig quickly or slowly LIE sliown in Figure 8. For ~ ? j r lfilter.5. The sensor signal ruils througll a bank yonds to a ctifTerent ( of vehicle using I Irr t e q l a t e s in F'igure 8. I)roblcrris of this kind oftcn occur.32.31.hc sigrml .33.t ) Example 8. Typical foriris of signal for. ancl h s e events arc automaticnlly couiited to give dehirrd result.:Y.k Coiantmg and rnrastning tkip speed of vehicles .t. Thc only soil of information a \ d i \ b l ~ is t h c signal from tkie erisor.
Convolution &rid Inipulse Respoiise Slow car I Fast car Slow lorry J'igure 8. The impulse rcspous~of a room call by heard as the response to a clap or the crack of a d r i p . be clone in reverse to rernove t2re interference.33: Typical signal f o r m Figure 8 34.I90 8. This convolution must. . Block di'tgiatn ol the detector A microphonc recording made in a room is the convohitiorr of the sound sigiial produced by the iiistrumcnt or voice with the impiilsc response of the surrounding space.
cribed by iui impulse Chwter 8.TIsy.4) for To + U. Excwises 191 e A blurred image horn a camera car1 be ixiterpre siiiothirig system Thr irnpiilw rCspCJlmc is t hc blurred spot Lhat would be a sharp poiiil on 12ic original image A radio cliannel with multipath propagation can be clc. because the mcasurcii signals usually have additional iioisc interkrence.8. then it is porsible 1 0 atterript to invert the iirfliiencc of I ) ( i ) by convolution with i t second J.erii h c tiou). The filter that rcpresents the best coriipronrisc betwern decoiivolitioii and noise reduction will be introduced in ptcr I8 as the Wiencr filter.4.tcin g ( 0 . T .' h ( ~ ) c l t . using (lie cmwolution theorem: (8. (8.77) The systcm function G ( s ) = L ( g ( t ) }of the second sy5tern c'ari b t i cxpiessed by the systeixr function H ( s ) = L { h ( t ) } .78). (*annotbe ~ccovered froni the second system. It is he rarely possible t o carry out in the ideal form ot (8. Iii the icleal case the result is agaiii a delta irrrpulsc: S(t) A h j t ) = 6 ( t ) .78) 1 effect of a systnn with system function G(s) is called deconvoZut/on. determine: 3L a) f = J' e . This problem will be dealt with in Scctioii 16.1 Show that the response of an RCcirmiL to a short rectangle impnlsc (8.6. Exorcise 8.d by ixit RCkircuit with a largc. Exercise 8.. sigrial compoireiits that weie suppressed completely by thc first sytitein (ZP of thc syst.78) could also be unstable. A sy rri with systcin ftirrction G ( s ) (8.4 ) . The irrqulsc ~ n cmi offeii be rnotlrl1t. s If the i~iipulwrespoiise h ( t ) is known tor sucli a sybtcrn. Jri aclditioa. 3 response that consists of multiple timeshift cd echos ( The inertia of ine~~stiring instruments has a smoothing effect on the measiir~cfsigrialb aiitl srnooths sharp stcps into slowly rising slopes.7) turns into the irnpiilse response (8.3.2 Using the calculation rules for thr delta impulse.
xereise 8.t ) xercise 8. te I cx S( 1.2tj dt 0 2 4 6 ’ Giw .rt7(t)iisirrg t.5 Form the derivation of .r. Convolutioii and Impulse Resporise 4 3c fd ‘ = .hc stcp function E ( t ) for Irmtli.?(t) and . Form the derivation of j’(t) = ~ ( .192 8. ~ X ~ 8.6 ~ ~ C ~ S ~ You a ~ giTerr the fuiictions f ( t ) = t e ( t ) and g ( t ) = ~ ( t )E ( [ .7 Esairriire the followixig signals: . Calculate e y(t) = f ( t ) * g(t) iiriiig the conwohition intcygal. f ( t )= ~ ( n t ) .4). t h m f o r m tlie derivations aiitl sketch them xercise 8.
iliai Exercise 8.3.xiable) . giw a) the systrrri function (note that r(t) is the output \.3).8 Find the responsr of thc HCnctwor k shown in Figure 8. Note: see Srctiorr 8.2. xercise 8. created thiv result. cleterrnine the pair of sigrrttls c.I ( 1 x  r )d r .4.3 h y solviiig the coiivolrit ion iiiiegrd f h ( z ) .t t 1 ~ 0 1 2 I t t :Lid dJ For tia(li of the convolution proclirts.1 to the reclarrglp impulse. For t h e example ot an RLC"net\vork (Chapter 3. 0 0 show11 izi Figure 8.
50 J that the system response ~ ( t ) coiiverges? Assunie that ~ ( tis) right sided.1)) the irnpdse r erponse c) the region or conveigence o i the system function cl) W2nc*hvdiici t a n the paranietcrs 00 and w ~ of the input signal u ( t ) take. .
3 We will ltarri furthei impou. Attempting to describe such LTIsystems using the Laplace transform leads to s in fiinctions with an iiifiriite i~itmhei poles and of Ycros. the Foixier transform plays an iniportarii role in iid tlie rriodellirig of LTIsystems. propertics and iisage will be examined in this chapter. The bailie also gc L{coswt} arid C{l}.is lost.aplace traiisform as oveilqpirig eigrnfiinctiims o LTIsy stern. It allotvcd us to evaluate he svstern re5ponse 111 an elegant way. or uiiilatei a1 signals.c arc' niore aspc s of oui prcviotis dealings with tlie Laplace trarisfoi in that we have riot yet iriatfe fully c l e w putting the systerii iriodel in the freqiicncy ..4. In this chaptei \vc will see how elegczntly t o I i w s this problem if generalised funclions in t hr frcqiic. or with chsential singi ties.tp systems that were described hy cliffereiitial eqiiations. although L { P ' } docs iiot exist. We have represerrttd these signals wit fir the 1. 'Yhc laplacr transforiii i s pcvr Iicularly suitablc it tve arc concerned with signals of' finite diiration.4. The olwio tages of the Laplace t r arisform only conic wit 1 LTIsystems that art3 ( h i by a (*lclitrnrmiber of and Leros.ncycio~iiaiii are permitted. hiit the iiia niatically correct methods require sigiiiticantlg more advaiiced knowledge of funclion t lieoiy than we have discussed in Chaptcxr 5 Even worse is thr fact that the clcgancc of t2ic iiietliod riiodelling system ils a few complex natural Ieboiiariws in the timedomain arid fretiiienc3/rlorKiain .e transform. It is act1iillly possible to work arouird these diflicvlt ies. orie may ask why there is a need for anotlirr transforni. T1irr. which hove f the h i m t". in particular if thew werc initial cwnditioris to consider. Its defiiition. After WP haw already provcn that ihci Laplace transforiii is very useliil. tant exarnples in this chapter. that cannot be characterised by M few polcs and ~ e r o sfor exaiiiplr the delay ciicuit iii C h y t c i 8. We will th chapler with a c*riticaLreview of Laplace transform. r e a clescribecl by reiitial equatioii with coi coc~ffkients Other iniportarit LTIsystems also exist. tlricb evie ace sfor w~ got to kiiow the Laplace t1aiisforrn In ll1e ~'LeVlollsch. howevc~r. 01 eqiiival~ritly.
oiilv pmvides a simple bc> dcscrihed by o r t i i n q ~ diI The l. but 111 tlre tiinr.domain licts the prirntlrv goal of making ( ill(~l1lftti(>ns siiiiplty to cairy o i i ~than I hey ~ v o ~ be in the tirnedomain.trit cocflicie~h.tplace transloorm tlc fuiirtioris of Ilrlsvsteins. while also Iwepirrg m m y of the riicrits. rf ilc1. ntl ex. 1 J lie iriterprct at i o n of w c l i polczero plots is riot so siriiple.211 returrretl to the timedomain after rrsirrg the advintages LVcA l w . that axe eigerrrrr 1x1 thr Srcquencv domain systein fuiictiorr. h o ~ c ~ / ebtcarise the r. xist tor sigiials cst. that c m 1 ations with ( onstnrit coeilicicwts. which rim he wprcsentccl in tire fiequeric ydor~i.. In .3). wc' iiow cwiisrd(>rt2iv Fouricy transthat it owrcornes thc stated disadvm?tagesof t he L ~ p l a c r transf'orrn. hi)wcm~r. It fails completclv eiitiai e q u a \ b n with cmist.tittit(i a miiiiltiplication by the bystcwi fii n.vile‘ have (. dc splanr is diffictilt.iy circuit. The first c*xarriple of this i s the evaluation ld t m response. example.cloinain rcqiiirps 1he uscl ol the corivohition integral. ie . The oirlj t i m e bring the description of P 1 1fiiiictiori by the poles and zeros in the cornplcx splane (sw C'hapter 6. (he piopclit ies of a As an alteriiative to the Laplace transform. ~ rarcl) at tempted t o clrai a c t w i s c or dehign roperties in L ~ frecluencycloMiaiii.ornp1~x fnnction oi to1 variablPs. As a. E101~ thr 1 1 iriotlel for UI'Isystcxrrc. VEV giw the arrdysis in natural resonancci whicli is (io siinplc ie frequericytlomniu using partial fractions. fbr H ( s ) has t o be niialytic in the region of ronvcigcrice. systeiri function i\ a c. The following point siiriiniar tage5 of the Laplace tiaiisform r l Ii. as this property mav be violated.hat we limv nwer doire it iii the tirriedomain.tinplc.
aiicl tlrertl are iii far t tlirio fiiiict ions 1 hilt although not iiitegrahlc. Igimicrit. The pltnsr .apl<ice. T'rrr argurnci j u s t a conveiition. symbol is rrol.ii11tl not S ( d ) and p(w). in order to g i w i i stricat matheniatic :tl rtit~airirig. i ) arid ~ ( J L .iginar>r axib is b&ig defined it could illso bc clefined as X(d) on thc real axis. ((3. S ~ J s p 7 p ( 7 w ) is usually c'upr d bp the phase p ( l w ) . strict nmt hcthe J.6) . ) .3).ioii of the Foiirier Transforiri 197 Its inagnitiiclc \ X ( J J ) i called the rnngrirfudc spectrum. have ii Fourier tramfoxin. to iase rliffcreiit synibols for diffrrerit t iaiisforriis.3) will br iised CT7c use [lie same relit transfornir.o eatli other in the original a i d trarisform clmiain'.2.tnsforiiis aie (9. Somr. and there arc actinally rcvcral boolis tliat tlu this . if a relationship between Fourier and lap lac^ tr ansforrris i s made (Sectioii 9.ion in the an mrnbcr plarre on t. tcxthoolis atteiript.tlie clcpeiiclenc*y OII iil is im iigpd. for example. At the hegiiirriiig it is CO that oiic writes X ( j . For the Fourici [xansforiii?the abbreviated corrrspoudency form 1 (9. Euamples of sii(*h time hrncl ions and their Fourier tr. Dcfinit.9. which mcans that a onediinrnrionnl cwniplrs tiinct. Tlic sciise of tlie ju! ronvrntion becoiiies irurnediai (sly c lcar.hc iiir.4) This con& ioii is sufficient but iiot i i ~ ary. but typographical shorthand for 'corr espoiiclixig t. Foiirier aid latcr.
Ro. Sonir eltmcntctry examples are con nts ( ~ .he trigovtornctric frmctions siriwi. thc extension to cover gewra1isc. h ~ Forrrier trarisforrn is defined on thc imaginary axis ancl not 0x1 the real axis. sitirit a5 i t s Lixplace transtoriii on the imagina That is of coiiise only triie if the r.aplacc transiorm cxists as well. The Fourier tr.t and cob wet.iou bctween Laplace and Fourier t ransl'orrns is: if the region of coriwrgerie<' of E { .tlml Fjzjt)} = C{:r(t)} j (9. ities a ee s S rie it can bc scen immediately that llie FoinieI transform of a function of tiinc is the axis s = j w of the coniplex plane.and the s i cp fuiict ion ~ ( t ) \lit\ will deal wit. T h e popularity of t hc Fouricr transform conies from the posqiblity of assigning a spcctnin7 with the help of the dcltit irripiilsc to riiariy iniporkd prattica. with Fourier traxisionns that arc fiuietions of a.r ( f ) } contains tkic imaginary axis s = 71*'.d functions does make smse.h the evaluation of tlic Foiirier tmniornis iii thc coursc of this chapter.tnsform of the s k p fimction contains a delta impitlse 6(d).wver.9) s=. TW ltax7e encounimeti it in the timetlorriain. g . r ( t ) = 1) arid t. biit not in the ~requenc. real parameter w (jiist as with fimci ioris of tirtci).21functions for which 1 hew is no Laplacc transiorm. to Up iiow. . there is no room there Erw gmrrafiscd functions. As Laplace trimsforms in the region of corivcrgeiice are analyt ic fimctions.7w Wit12 (9.9) the corivention sliould r i m 7 i~iakesense.ydorrii~in. t h a l t . T h e coimcct.
1 fir Fourier transkorrn (9.10) into the clefintion of thrx Pomicr txansforni (9.s 2 ( s .tiiis the imaginary axis. In practice.15) .10). hccauie the Fourier integral does r i d conveige.s bc+wecn Fourier a~td Laplace Transforms 199 Example 9. 1 yL+n (9. Lhis mclarrs that ?U: is replaced by s and the rcgiori of convergcuc.a.(t) = &(t)e? has tfre Laplace trarisfoum LI >0 (9.r 5. The Lttplace tiarisforni can tkierciorr easily be obtained from tlie Fourirr tr.3).t. If the rcgiori of c onvergence of the Lapldce tr arisforni cwiitains the irnaginai y axis.jw)2 = I 1 X(s) = = 1.rigd (9.13) has rio singularities for rcal values of w aid c m be diffcrent i a l d at all points.9) can thcrr be differentiated any riiirrihcr of 1 imes.14) + 1) . the Fomier transforin tlors riot exist. dcfiiics the Laplace t iansform Example 9. although the region of coiivergcnc*clriow lies to the right of the positive mimtter a arid no longer corit. and tlius contains thc imaginary axis of the splane with Re{s} = 0.iows the coiinection (13.2 Wiat is tlie Laplace transform of the signal wil 11 Fourier traiisforrn X(.>ifF(irenc*t.12) This clearly sl.l( )s (9. 011 the othcr hand i s unch. however.9) betwreri the Foiirier and Laplace traiisforms If.3. iFTe can porform analytir conlimitition 0x1 X ( j w ) .11) Tlie region of ronvcqcncc lies i o tii(2 right of the negative rinmhcr . wlieie we write 1 x (.9.jU)= 1 I____ ? l+W‘L (9. The Laplace tritnsforni. Sirriilarities arid r.13) Expression (9.10) (9.jw) I .uisform using analytic tontinuation (C31aptt. (9.1 The functioii . The FoLrrirr transform is obtained by substituting (9. U < 0 irr (9.(.11).4.1) arid evaluating the iritcgial X(jw) = 7 .
We will also S(Y that signals likc s(t) = c I d z and others tlcrivcd fioiri it (constail s.can bc in1 directly as a computer program or a5 circuit. which e wc will rder to later. \%it11 it.200 9.Lin the imaginary axis. I properties can he re imriictliately fioni tlie Foui irr transform oi the syitem response. Figure 9.1 shoms the corrcspoiidiiig tr. as in this c&e the conditions for (9. The Foiirier transform also has great irnportaricbe f o r digital signd p r o its diwrcte couiiterpart. In Example 9.(~iiencyclomairi.trrsform pair t .1). the discr Fourier transform (DFT).The. This is particularly q q a r e r i t ers) that clrtaracter d hy theii t~ equrnry belravioiir. tlicwforc 1 < ne(s} < 1 .9) have been fulfillcd. trigoiionietric function%. etc.>ipliice tramforin. bint no Fouricr transform. t~isrretc’tirlle5iglials cd17 be w01Led with clirectly in the fic. The Fourier transform is thercfow a nccessnrj and mcwiirigful co~npaiiioriof t he l.) liavr a FouriPI transform. 1)ut no Laplaw transform. (9.1 we saw that signals exisl which have ik Lapliice tramforin.. 4 111 this svction w mill firmcl thc F’ourier transform of mine important signals. The Fourier traiisforrn o i the delta irripulse is obtained by inserting it into the Fouricr integral (9.16) ‘lhc result c~oirespontlsto the Laplace Ir m s f o r i r i . as the fLOC m i s t cont.1 arid s = t1. Foirrier Tritiisforrn Irns poles at X (A) s = . and rising its selertim propc’rty.
we first inlruduce an abbreviatrori.9 . arid deal wiLh the syrri1)ol r e d as t.he function definecl in (9.t(l) to indic. Wc thercforr iibe tlrc notation rcr. Exaniples of the Fouricr Transform JVc obtain the Fourier trarislorm of 201 s2iiftftCd delta iinpiilse the same way: i3 arg{ y{ 6(tz)}}= m Fourier To firicl t l i ~ Foiirier transform of the r( iigle function.3). 4 .18). (9. of which WP will make widrqtiead IM’. h i t dealing witli the piecewise tiefinition in (9.at.18) .cl thc rcctanglc. lirncticin. A I tangle fiincticxi is easy to dracv (see Figiirv 9.18) i\ not suitable b r t aiisc of the ~ic’ce ry ( asc distinction.
4. The Fourier Transform rect(t) 1 Figure 9.19) The Fouricr transform of the rectangle furiction is obLainet1 by evaluating the Fouricr integral (‘3.F{rert(t)} is ~liowii Figure 9.20) 7‘hr plot of tlic traiisforin . an abbreviation for this clraracteristic* function. > 0: rect(af) = i 1 0 for I1 5 t 2a otherwise 1 (9. We call .202 9. in Figure 9.4: Fourier trmsforai of the rectangle ftriiction ?VP will also intwduct.3: The r w t anglc Ftirictioii with 8 fafitor a.
22) riglc bttwwn the mairi peak m t l the first two Lero points t o t1w iiglit and left of 11= 0 (see Figure 9.6). we obtaiiied at llik point from the 1'Hopital rule. Exainplcs of the Fouricr Transform it the sifuiiction ancl dcfine it as 203 (9. we can w r y clegaiitly forrmilate the correspondence b e rween the rectangle lunction ancl i t s Fourier transform.0 corresponds to the liniit tliat. si(v) is continuous. Figure "3.5: Thc sifiinction For the entire area under the iiitegral (9. By roniparing (9. we find that (9.21). dq~eitdexiton tlic dimeriiionless vctriable U.23) . whether it is s c a l d in the liorixontal or the vertical axis With the sifimction.4. This rule is very l)ractical as i t can be also used for m y other sifunction.5 sliows tlie course of s i ( u ) .9.20) arid (9. Figure 9.21) 1 for v = 0 In spite ol the value z/ iii the dcnoimnator. so I/ .
Pigitrc R 6: The aie. roriesyontis t o the integral OS the sifiiti(. the Foiirier transform eJ'"Ot mii br givm in the form of a distribution.tiori (9.lrss. we consitler first of all a section of the oscillation . rrpresrntrti for clitrei erit :is the iiitegral that arises tlcnrly does iiot conwrge for w = do.7.r o f tlic triatiglt. spedra art. wr l i s p the sifnnction that we have just iritiotliic In o r clrr (0ov(aicorii(' the niciitioiicd convergence prol)lein for iritegratiori of an osciliation of infinite duratioii. To ( h i v e tliis. rcctarigle fimctioiis ancl thrir.22) I In Figure 9. Neverthc.
8 shows a.25) con of the signal of tiniie tluralioii . It is tlie same as the spectriini diowir in 7 . plot of this spectrum.t).r(t) 1s also obtAnt.tl from .Figure %7: Various rectaiiglc fiuict icmh rect(c2t) and tlicir P'oui iei transfoiins with finite durat$ion: (9. but i b shifted by W O on the frcqiieiicy axis. the cornplex expoiicvititil fiirirtion in (9.rr.(. with tIie limit T x Lli~ speTctrunr 01 .26) Figure 9.
do. X T ( . tlir same as I h e area under an isosceles triangle with its base between the two Lero points next to the niaxirnurn of the 4fimction (see Figure 9. (sec Figures 9. j The a w a iinder the sifunction is. can he expresqcd by a delta iinpiilse. aiid bo when the product X T ( . For large.8: Spectrum of a. so we ask whether it.JW) beconles a spike with height. that X.31) . l o r 7' + 00 (so a > 0). We form the prodiicl. of X r v ( j w ) aiid another function F ' j j w ) . that is constant at LC: =. That rrieaiis heie from Figlire 9.29) thiis becomes (9.206 9.8). The Foiirier Traiisforiii Figure 9.8 that. T ~ ) can clearly not be represciitetl by a n ordinary function. ~ ) F ( J for T J) oc! is integrated.28) we recognise from Figiir 7. complex rxponential function ol frnile duration (9. however. approaching infinity and width approaching wro. Equad ion (9. 'rO do this we must test XT(7ul) lor the selective propwty of the shifted delta impulse (8.7 arid 9.15) with the h i t Y' i m. j whcconies ) ever smaller (apart from at W O ) .ion. but can othcrwise bc any ftirict.(. values of T .6). orily the UP F ( . the fiirictioa X ~ ( . p o ) is tontribulc>d.
as lortg a b thew two limits c'ari bc cvaluatecl.17 ) : (9. We can tlicwfore writr 1 & 3") I (9* 3 2 ) Cornbiuing t hcsc: rcsnlts yields TVc have now ohtained thc transform pair for tlic complex expoiient ial function: (9. Exaniples of the Foiirier Transform 207 Tlic fuiictioii liml irx. As i.4. It helps to split tlic.36) s1iowir in giaphical form in Figwe 9.9: Temporal behavioiir of the fiirictioii f beyond limit.34 Xi.hc time funrtioii 3t t = 0 grows Figure 9. with 0 < E < T < W. is similar to thci transform pair for an impulse shiftud in t. From the iesiilt s of these two integrals. it is not to integrate over this poiiit.mplc a tiine fuiictioii whose spet%rumcaimol be found by siinplr for cvalrratiun of the Fourier integral is the function x(t) =  1 t II (9.35) ~~S~~~~~ of 1 7 A further ~xll.pJ) except at w = JO. Tliis procedure is also kiiowrr as cxlculating tlie $ . X 2 (id) clcarly rciiiows all valiws of F'(. The upper integxal is solvcd firqt for the finite limits t aid Z'.9.9. and thc lower iiitegral is the11 solvcd correspondingly foor 7' and t.1i(Jtiincvlomain (9. tlic solution tm be obtaiiitld I t y the two limits T 4 'x) and E 0. That is exactly the property (hat ithtifics the delta irnpulse. iiitegral into t w o parts for t < 0 wid 1 > 0 .
5 . The Fourier Traxrsforiri To e d t l a t e this irnpropr integral.208 9.4). although i t s Fourier transform caii be d e b o ~ again.utlr of thc sifunction.6). Please note that R ( gt' of sign of w also lcwls to a change of sign of the amplif.i*aluefor both positive and ncigat ive v a l i m of' frc~liierrcy (sec Figurr 9.r(t) = cannot b~ abwlutely iriteguated (9. e lliai. Becaiw of tlre disconLJ tiniiitv. we can i i i e (9. tliat if il function can he iiitegratc~d. is sary colidition for the exislerice of t h Foinier transform. X ( j u i ) caiinot be analytically c*ontiiiiicd Thc fiinrtion . or even beller the triaiiglc ride (Figure 9. Foi ij 0 the citlculation is trivial. We obtain 7 A. pitrely imagiiiarv spectrum is obtaiwtl that liar a constitrit .10).6).
+ t ) ) ) 1 2 .( ~ . ( t. ’Jkansforni 209 (9.41) Ever).43) .9.r<.5.(t) = . Svrnnretrirs of tlw ti’ourici. .tion can be split iirto a n cven arid a 1 odd 1 part where I ire t ernis are given by (9. func. (!I 44) T l i c dcfiriitirrris toi FVPII wltd odd functions arc equally valid for both real arrti coinplcx signals.N) (11.
ingly. a. f ~ manifest thenise!vcls iii 1 tie frryiic. there aye often different symmetsriesthat can he siimrnasisetl by the concept of conjugal. x ( t ) . X ( J W j = X*(.r(t) = Re{.jw) . By splitting . one after the other will apply both opcratio . it is ixnniodiately clear that.ry. ior n spectrum wilh conjugate ~yinri~ctry.48j es We will now use the synirrietry r r l a t i o d introduced to debcribe the conwith mlvaliw t i m e si nection betwwn lime signals and spectra. Here :c* ( t ) mearis the coirqde:~ ccinyqute of the jimctiori.e symme(.210 9. (9. furict ion with c.ple: functtort x( t ) has conjugate symrrtetq uihen: x ( t ) = x*(t) .ncgclooulclin. We s irivestigate how these pr )ciated wit11 corijiigate symmetry.orijngatr hymrrietr y RH art even hinctiorr real part and an odcl function as the imaginary part. Of c o i i ~ w tliih dcfiuition also applitxs to functions of frcqiitrtcy. Corrcspoiid.r(t)} + 31m{~(t)} into real and irnagiiiary parts. The Fomier Bansform For tthe yea1 and irnagiiiary parts of complex signals. A corn.
9. Re{eJ"'} = cos(w. Starting with thc real part of the spectrum. bearing in mind that. it According to Delixiitioii 13. IX(Jw)I = IX(JJI\ magnitudc cven.cvould further like to know which parts of the time signal corrtqmiid to tlie real mid imaginary parts oi the spectrurn.50) (9. Real signals ifiiis have corijjugate syinmctrical spectra (9. (9.5.1) for a real function of time ~ ( t ) : 211 X(.r (f) = x(f). All right sitfcs arc identical.I * (I). or all wnatively as rriagriitude and pliasc. Sgriillletries of the Fourirr 'Urti~isform t o the Foilvier int.egral (9.5$} (9. arg{X(jwj) = arg {X(jwj} phase odd.7~)} = [:(r)cos(wr)dr Re{X(p)}  l_?(i) x c o s ( d ) df . with the Fo1nic.i integral (9 1) we substkute z = t for the time variahlc. so for a real function of time with a l e d spt~t%nirri imist bc trne lhcxt . the operations c a r i d each other out.values of imaginary arid real paits.49) As spectra mith complex values can be represeiited real .jw) = [3t!C7i~t d M y( jw j := As the real func*tionx ( t ) does riot change wheii tiansforrrred to the coniplcx coiijugate func tiori . is . Im{X(jd)} = Irn{X(jd)} imagiiiary pait odd.54) 1%'~ . ~ ( t )an cvcn function. I n the last row we liaxc a g a i n sct 7 = 1. coiijriga X ( p ) can be expressed by even UI odd syrrinictry of thew coinporierits AY(gwj = X * ( . as the d w of thc integral does mt depcnd on the notation used for the iiitqyation variables.] w ) Re{X(p)} = Re{X(jw)) r t d part cven.51) (9 5 2 ) (9.p)} = [ f ( t ) cos(li?f)dt 3L Re{x(.t) is an even lunctlon: x Re{X(.
Pwn (‘3.5 liiis a11 imaginary and odd spectmrn. evrri c )  X ( .58) ry of t . be which brings US to the coiiclusion that a r d f h c t i o n of time with cm imagiuary spectrim1 iiiust bc an odd hmction. (9.ocid. Symmetry of Complex Signals ‘Tlic previous results for rwl arid for purely imaginary signals can now be uriittd t o give the symmetry for general complex signals.212 9.(tj}. otld. ‘X’hc.60) The imaginary fuiictiori s(!) ancl its real iniagirrsry part Ini{%r(t)} should not bc confriscd.E.56) Tliese general principles .5.56) aPc .4 The rectangle fiinction is real a i d eveii.i) real. j u ) imaginary. ~ (9.pectrum. evw X ( j w ‘ ) imagiiiary. Fourier Traiisforin The same dcductioiis ri~n wriade for the imaginary part of thp spectrurn.S). Every signal c m be split into its even and odd parts. Tlic equivalent rclatiorisliips t o (!J. e\wi sn(t) real.we ~ ~ c a l byd o u r observations ot the tranrforms of e certairi signals.59) x < > ( firrmginary. We have now sliowri the following rywnietry relationships for real signals: r.ci S(. xample 9. (9.5 The rpal am1 odd fixnctioii of lirric 5 ~ X 9. For prircly imaginary functioiis x ( t ) = .lh{x.7jj real. ‘The m i n e is true for i t s .4 S y ~ ~ ~ eImaginary Signals The same deductions for time signals with real valws can also be made for irnagiiiary time signals. odd ci . nxicl its real anti imtginary pitrts. octd ) ++ X(j. (955) (9. aiitl . (9. { t )iinaginary.(t) real..
Iwerse Fourier Transform 213 four tcrrns in totsal for the time signal.61) that changing the sign of the iiiiagiriary part in thc timedomain lias the following efkct in Lhe freciiiencydomain: F{a*(L)} Re(X.6 A complex signal ~ ( thas t h e Foirier trimsform X(ju).. . logical. (9.1). ( . Example 9.. a.56) for real signals can be iisetl for thc real part. and easy to remeniber. more concisely:  Using t .60) for imaginary sigiials can be used for the imaginary part.6. The sclieriie (9.jIrn{X.(jw)} .9. tlie gener it1 complex case is sinpisingly simplc. in fact. h syrrirrietry of tsbceven ancl odcl parts this c:a8nba wri F{.What is 3{ * ( t ) } ? ) x Tt can be tsakeIifrom (9. (9.(jw)} = .(jw)} . lot of similarity with thr. and (9.55)..61) can be formed from these.59). Foirrier integral (9.Re(X. Although the syiiimetries seem to get more and more complicatled. The symmetry relationships (9.c*(t)} = Kc{X. for tlie symmetry betwccn real and imaginary parts of the even aid odd parts of a timr signal arid spcctrum 1191. and also the spectrum.jIm{Xe(jw)} + jIni{Xf)(jw)). j ~ ) } Re{&.61 yields tbc' important transforni pair The iriverse of the Foiirier transform is an iiitcgral expression that ha:.w} TnX(g) Equation (9.(jw)) = X*(jw) * + gr(.
'Vhc significant diffeimces between the t ransfu~urma i d its inverse are thc intcgrat ion variables e ireqiirncy). the Fouricr as 7 transfoiml has many propelties that need t o be known to take advantage of the . if discontiriuities are foresccii and clealt with (Chnpter 4. Willi tlir selectivr propcrtv.igiiiary axis s = J W has been chosen as the path of integrntion. rie sfor As ~ ~ f l l lthe symiiietry properties that me have already dealt with.62) in accordance with (9.2). .62) call br interpn\tcd as an invcrse 1.1). this degree of unarnbiguoiisriess i s sufficient.4.62) leads .9) are fulfilld.1)) . Deviations at individual points of cliscontiniiities cio not change tlie value of the intcgrals in (9.4. to a function of time from which (with (9. the sign in front of tlie rxponeiitinl ftmction (/+) arid lie I 1/2n before the intrgral. which is tlie case for normal convergence of intc Just as foi the Ilaplece transfoim. tlie irnent of a time signal to its Fourier transform is irnarnbiguoiIs. (9. Lilw thr iriveise Laplace transform. but in this c a i e only in~tlampcdostillat ions are $1 errriitt d . This iritrrprctatiori is valid if thc conditions fox X(s) = X ( J W 43w (9. the inverse Fourier t r ansforin can rcprcsent a suyeriiripositiori of cigcnfinictioiis r J w tof an LTlsystelrl. )~ the region of convcrgeilce [or X ( s ) encloscs the iniagiriary axis arid X ( ~ J ) hc can differentiated any nurnbcr of times. put tlle defirrition of X(p) we irito (9. and exchange the sequence o T integrations.md whcii solving p r a c t h l prohlerns. p ) c5tli be cal(~ulatec1.21 4 9. b(t = 7) s(i) (9.6.3.aplncc trarlsfornl (5. Irk fac*t.34) for which the irn. Thcrc i s an exprrwSon (ill square braxkets) within the outcr integral that we can iiiterpiei (see Section 9. X ( .3) as a clelta impulse shifted in the timedorniiin. we cwd up with r ( f ) Tlius we have s e w that thc inttgral in (9. and trcquency pararn ~ real valurs. the roles of ui arid t have been swapped. The Fourier Transform The c m s of this similarity i s that both the tinit.62) actua11y yidds the corresponding furiction oi iiriic for a possibly discontimious spectrum S( p).63) This der ivatiori requires that both improper integrals can be swapped.63). In contrast to Section 9. To vt?rify that (9.
liaviirg to sohc integrals over and over agaiir. a correspondence l ~ t w e c n the frequencydomain and timcdomairi c m be derived by ci\rrflllly choosing the argiintents of a given transforni pair. Example 9. for the transform and its invrise. h. Properties of the Fourier Transform ~ 215 frecluenc.66) This relationship is called cludzly. With appropriatc subtitutions into the integralz.1 ~ i ~ e ~ r the~ y of i It follows dirertlj from tlie linearity of‘ t h c intcgratiori that the pririciplr of snperpositioii applies to the Fouiier transform and its inversc: (6.ydomaiii representatiorr.12 can be obtaincd as the sum of the transforms of two individual impulses that ale eiLch shifted hy f r : 6 ( t + z ) + ~ ( t r) o.7.12: The inipiilbe pair arid its Fourier transform Hccausc of the similarity betwccii the formulae for the Fourier transform ancl its inverse.ants. 9. r and d c m be m y real or complex ronst. (9.7 Tlre Fourier trimsforni of a delta impulse pair as showii in Figure 9. . withiit. e’wz+e’d‘=2coswz.65) Figure 9.7. it (’a11 bc shown that for two functions J 1 and f z that (9.11.
23) that we already know as the traiisforiri pair for a rectangleshaped spectrnrn =+ si( i) 0 0  2nrect(w) = 2xrect(d) . This is because tliere is no analytical coiitirluation into the cornplcx splane for the dcltn impulses on the irnagirrary axis.67) Figure 9.ioii Exarnple 9. .r. For sucli fitnctions of time there is a Fourier Lransforrn. L.y of thc sifimction and rectangle fimct.9 Fhni (9. 'The Foxtrier Transform Example 9.8 Using the property of duality we call obt.68) T h Fourier t. but no Laplace transform.ain from the relationship (9.216 9. (9.13 shows the duality between the sifunction arid the rectangle fimctiorr L. Figiirc 9.13: Dualii.65) WP obtain using the duality property of the Fomier transform of a sinuhoidal function of tirrw: cosw()t 00  7T [ S ( d + WO) + S(d  d) o] I (9.\nsforrri consists of a pair of delta irnpulses.
14 shows the rclationsliip bet wee11 convolution with tlir iiripulse resporw arid rriiiltiplicatioii with the frcqueiicy response. we hnd the Fourier traiisforiii of t l i ~ triaiigiilar impulse shown in Figwe 9. (9. but it is somewhal tedious.7. ~ ~ ~~ ~~ ai ~ ~ ~ o t~ ye ~ We s a w wheii dealing with rertmglc fuuctions that thin rcrtanglc.F{x(t)} with the Fourier integral is incteed not so difficult. It is much simpler tTonoticc that .70) The proof is carried out irr much the same way as for tlir convolution proycrty for ti airsform (romparc Chnpt. This is also trite for all fiinrtions of time and tlieir corresporrdiug spertra. and vice VPL The fartor n may hc npgativc.15.er 8.. arid vice versa (IGgure 9. The most iruportarit use i s thc calculation of the output sigiials of L'I'lsvstterns.2). The sirbstitiition 1' = at :md U' = U / ( I in thr intcgrals of the transform arid irivcwcx transform show tliut cornpressing the Lime axis corresponds t o t2xp. Figwe 9.10 As ari example for the use of the convolution property.71) Chlculnting the E'ourirr transfolm X ( j w ) = . Tlir Fourier transforizis of the inipiilsc rcspoiise is called the frequency wsyonse.9. wliicli is defined as i~ function of tiinc by (9. but not complex: (9.7).iiitliiig thc frcyuency axis.69) This tkicorem 11m an importan( impact on tlie socded 'timPbaiidwidtIri product'. which will be tliscusscd in Section 9.10.9.4. Example 9. have wide slmtra. Properties of the Fourier Transform 217 9.
73) follows immediately.72) I I I (9.1 6: Triangular impulse as yieldccl t ~ ylie corivolutioir of two rectangular impulses t also Itt.15: Triarigular impulse a triangular impnlsc is the result of tlie convohiticm of ~ W J rectangidar impulses.113 this can he easily coniirnied by inspection.he corivolLiCion property. Fonrier transform .14: Thr relatioirship between corivoliilion a i d multiplicatioil Figure 9. expressed by (9. rect(t) 0 * rect(t) 0 0 = :r(f) 0 (9.72). The Fourier Transforrn Figure 9. Thc. (9. With Figure 9. This connectioii can Figure 9.73) With t.218 9.
s ~ ~ ~ t i ~ ~ i c a t i ~ n Ry using the duality o f . As the triangular inipiilsc is real and even. Thc signal can only be observed foi n lirnilecl . its spectrum has the same properties because of tlic.7. i 4z 2z f i Figure 9. wc obtain from the convolution property.74) Figure 9.17: Fourier transform o t.he triangular impulse E 9.7. Properties of the Fourier Transform we wcm looking for is already there: 219 (9.t7 shows i t s plot.Tl (9.66). I 0 I (9. symmetry relationships. The spectnmr of two supeiimposetl siriiisoidal signals with almost the Tame freque~icy to b~ rcpresciited iii such a way Ihat Imth signals can be easily clistinis guisbPtl in the hequeiic.11 As a more t l e t d e d exitinple. Example 9. F the relationship for the Fourier trxisfonii of a signal y(t). rind .that can be written as the product of two signals f ( t ) arid y(t).9. wc consider a Lypieal sp ral analysis problem.75) as coiivolutiori of f hc spwtra F ( j w ) and G( / U ) .ydornaiii.
The fimction tlescribcs tlrc finite window of observation through which WP sw thc sinusoidal function.220 9. Figure :I.79) .he two signals from each other.yA(t) and 1 hc window of observation S ( t ) W e already know that the spectruiri of a siiiusoidal signal corrsists of two delta irnpulbcs o i r the frequerrcy axis. The location of the irnpiilses corresponds to l. the delta impulses will lie close to eat41 othcr: g ( t ) = .fJ(i) = f i t ) .hr frequency of the signal.7 ‘ / 2 I o T. The obselved signal i s t ha (9. 1 2n F(:jw)* G(:J’hi) (9.19.76) The firiitr observation time can be exprrssrd inathernatically a mn~~ltipliral. b r ( t ) + gz(t)l ? as shown in Figure 9.ioIi s hy a rectarrgle function in the tiincdomain. P’igitre 9. Y(jW) = . As both frequencies arc nlmost the sailits. (9.’2.18 shmw the two sigiinlci whose superiiiipositioii we will be ohserving froin time . a the convolution of the spectrum G ( j w ) wliicli contains s t he itlral delti1irnpulsc>with the spectrtini of the window of observation.78) A4 = f ( t ). nncl we wish to determine for how long it needs to be olrxc~vd u s for t o able to distitiguish t. Tlie spectrurn i~(*tually ol~servetl m be rrpr t the multiplication rule. Because of tlie convuliition wilh foul shifted delta impulses. ’I‘he Foririer Trarisforrn  duration. 18: Siriusoidal sigrids 9 1 ( t ) .9l(t) + gz(t) = c o s d l t + coswzt 0 I (9.lready know the spectrum of the window: a sifunction. the spect r i m of the measured sigriat contains f o i terms that each have the smic form as the spectrum of’ tlie ~ window of observation and are located a t the signal flequency.77) ?Ye a.
a n d appear as a single signal.tr ) : (9.c) = ~ (. A window of observation with lciigthr T = 1. m7e obtain the sh.19: Signal y ( t ) weighted with the observation window tsi (wwz) Figure 9. Frorn the similarity property we know that the widei the wiridow. the thiriiicr the spectrum of the window.c) in the convolution theorem (9.70). but are still not recogniseable as two sqmrate irnpiilses. ~ (.9.21 (top).6 Shift and If we set one convolution partner equal to d(t .20 s11oivs the spectrum of the reference signal for three different widths of the wiiidovv of observation. can the two signals be clearly identified. The two c o m p i e n t s first staxt to become distinguishable at a window width of T = 1. ) The Fmrier transform of ~ (.2 ha5 a spectrum wide enough so that the components o both frequencies can no longer be f distinguished from eadi othei.80) Example 9. that means the longer f the duration o measurement. Properties of the Fourier Transform 221 Figure 9.12 In Figure 9.?ft tlz~orrtn with x ( t ) * 6(t .5.7. Only by significantly increasing the duration of nieasiirernenf.7. the equation r ( t )= si(nl)o e X ( ~ w= r e d ( % ) is shown. to T = 10.5) can be giver1 directly by the shift t theorem: (081 ~ . for example. At leash several periods of lhe frequency difference betmwi the two signals must be observed for them to hc spectrally separated i 3 +si ( w + w z )  :)J 9.t1) and .
A &iXt in the fi.222 9.)).82) Both pairs arc likewise illustrated in Figure 9.ial fiiriction is known as moduZa.83) wlieie the miiltiplieation of a time signal with R complex rxpoimt. as the spectrim is ori1y rnultiplied b y a complex exponential hincBtion. The above example sliows an interesting property of the shift theorem: shifting the signal does not change the magnitude spectrum.84) . re&( ). Thr Fouricr Traiisform and .  X ( j ( w .75) for multiplication with eJW0f 00  k h ( W .tion.ydorrrttirltlescribcd by thc m o f ~ theorem ~ ~ ~ o ~ is ~ ~ eJ”%(t) 0. (9.e(~~ienc. 2n W (9.21 (cerlti e aiid bottom). Tlie niotliilation theorem is a special case of the iidtiplication theorem (9.WO)) (9.W.
9.21: Example of thc shift theorern f c 5 ' Figure 9.13 .7. Proporties of the Fortrier '1Yansforin 223 t Vigure 9.22: An rxainple of the inodulatioii i.hrore~rl Example 9.
7): (9.o The cliRererltiation of the Fourier spectrum corresponds as with the Laplace transform to a multiplication of‘ the time signal with t (compare Chapter 4. t. Differentiation in the timeciornain col responds to convolut.86) As might be expected.88) cu . Note that. so its real part is represented by the solid line ().jwO = j corresponds to shifting the spectrum by 4 W Q = 4 t o the riglit. This corresponds to a convolution with tlie unit step fiinction I I‘ 2( z)d7 r ( t )* E ( t ) .he bilateral Laplace transform (comparc Chapter 4. ifkrentiation Theorem A s with the bilateral Laplace transform.1) with u respect to 3 (see Exercise (3.7.7. The Fourier Transform Figure 9. and its irnagiiiary part by the dotted line (). 9.4) by substituting s with :jw.22 illustrates the eKecL of InotiiiIating the signal r ( i ) = si(t) with e’“‘. ~ ~ t ~ ~ rTheorem ~ i a ~ i o Here we are only irittrestecl in inttAgr ating ill the timedomain.85) (9. there is a differentiation theorem for both the timedomain and frequencydomain. it must be possihlc t o diffcrt. can be derived from the dilferentiation theorem of t.7. This requires ~ ( t ) be differentiable. Nole t h t a modulation with frecpiicy . ‘1‘he time signal is complex after the modulation.14). (9. it.87) This theorem is therefore also called ~ r ~ ~ u l t ~ ~ l zwitht ~ o It can easily bc de~none a t’ n strated by difterentiatiiig the defining eyiiatition of the Fourier transform (9.224 9.titiate X’(1W).ioir wit81i&(t): = 2 ( i ) * S ( f ) dx(t) dt (‘3.
In fact.75) arid writc out the Foiiricr integral for ttic fuiwtions of tiirtc and the convolution in clet ail (9. T  A turt2iei propc~iy the Fouricir transform expessc.91) + 7rX(O)& ( U ) .89) 3 arid the odd tcmi 2s 1 i:i =7rh(u) (9.27) is not part of thc region of convergence of the Laplace traIisiomi.8. To derive this theorem we htart with the multiplication thcwreirr (9. Parseval's Theorem 225 c(/) 1  1 2 1 + iign(t).9.27).94) . It says ot that the integral of t h p prodnrt of two iurictions of time can also be expressed ah thc iiitegral of thc product of their spertra. 0 for signals with a iio1ixc~io mean. Pa~srvaZ'stheorem. (9. transform (4.93) ran therefore not he analytically coiitinuccl sincc it contaiiis a delt. as integratiori generates cz polr iLt B = 0. tfie i giiiary axis according t o (4. w 'Thc form thus oMained is incomplete mid i s only valid for signals with zero mean X(0) (1. 2 (9.a impulse at w =. (9.90) (9.93) The iiitegration theorem of t h e Fourier traizsfonii (*annotbe obLairiec1 sirnply by inserting 5 = j hito tlic integration theorem of the Laplarc.
p j is (9.97) For g ( t ) = S(t).ti permits the use ol coniplex xigiids.. We can rmd from the SyJklnictry property (9.q(t) we will also perinii corriplex fimctions. of the voltage IJhirig the rriagnit iidc sqtiart.226 9. Parsevnl’s theorem says that tlie energy of a t i m e signal can be calculated not only in the tiriic~tlomai~i.7 that for a signal g ( t ) arid the correspondiug complex corijiigate signal g*(1) t h a t the relationship betwt.grating the hut rncignitude squared specdrum. also in the frequencytiori~aiu.by iritc.61) a i d Examplc 9.96) Froni liere we ran obtain the geiirral form of Parsed’s theorem for coiiiplex funciions of t h i e (9.cn tlie spcywa G ( p ) arid G * ( .f ( t ) is gzven by T h e reasori that this integral is called the energy of a sigiial betonic~sclear when WP imagine j ( t ) .98) x I To iiilerpret this formula wr define Lhe energy of a tirrw signal. we obtain the easy t o rcniember relationship (9. . for example. The ciiergv converted to heat in the resistance is proportional to the iiitegral over tlir sqnarc. Only the magnitude is irivolwtl. to be a decreasing voltage acioss an ohmic rrsistaircnc. tlie expoiitkntial term ail tl 0x1 For the fiirictions of time S ( t j and . 15: Energy of a time signal The linergy E f of n siqnnl .95) For w = 0. The F’ourier Transform the left side of llie Fouric>rintegral disappears (9. thr phase clearly has no effect on the eriergv of the signal.
(9. Correlation of Deterministic Sigmls 227 Example 9.cgrating the rcct angle function: We lriave iised the fact t h a t tlir anglc.119) directly in the timcdomain i s possible. rfNTeconsider the product of f i t ) and y ( t ) ..100) and Parseval’s throrern wc can calcinlate the ericrgy much more easily in tllc ~req~~cnc. So fa1 NT have not encountered iandom sigrtals. where it. This lead. Combining two functions of time f ( L ) iind y(f) can bc more generally forntuii in (9. we will calculate the eiiergy & of t lie signal (9. but requires sonie tricks and rearrangerneiits.67) (9.14 As an example of the use of Parseval’s theorem. shifted in time by z. will play a central role. the integrd expiession formed is a fiinction of the separation iii tiinr 7.97).ydurn~ixi by irit. bVith the trarisftmrt pair. q u n l ~ of a rcctangle fuiictioii is again a rwt Tlie coiiccpt of c orrelatioii act iutlly originates from random sigiial t heory.9. to the definition of the crosscorrelation fuiiction . and we would like to erriphasisc that the followirig discussion is ti of ill1 liiriitrtl CO detcirriinistic (nonranctoin) signals.9. which we will iritrotiticr in Chapters 17 and 18.
(II.103) The crosscorrelation ltinetion tlescrik how two sigirals are relatrd whilc 1)cariiig in xniitd any possihle displacement. ~ We will come back to this later.1 Relationship with Gorivolutiori Thc integral cxprcssion in thc definition of crossroirrlalion (9.104) . The Fourier 'lkansfurin (9. The autocorrelation fimction shows how siiriilar the coinponrnts of a time sigiial D S P that O C ~ L atT diflcrcnt points in time. In fact.102) to (over raiitloni sigiids.z ) .)dL= f ( z ) * y * ( . biit the significant properties are already clear for cleterininistic signals. (9.102) is ccrt aiiily r&td io t l i ~ convolutioii iiitegral.102) x (9. when we haw ex~eucled(9. ~oper~ies 9.2.9.228 9.102) can be cliangt'd t o a convolutiori with the substitution t' = t The crosscorrelation EuucLion for determinist ic sigiiaIs cwi therefore also be defirted by a convolution I glfu(") = i f ( t > + r)g*(f.
*ir) L) * 2 ( L ) c .to) .L.ngingthe t ~ v o functions f ( t ) mid sit). =P 9.js z ( Z .z) .106) for the coinplex crosscorrrdation fiinctiori P f f W = Pjf(4 (9.ercha. pys(4 = = = Y(L) h( z) * I. (9.1.tiori function therefore has no general synimetry propertics. Correlal ion of Detaministic Sigiiwls Likmke for the autocorrelation fknction 229 (W Example 9.t O ) .l5 What is the crossc.P T L 1 z) 4 ( t) The crosscorrelalioii twtwen the output arid iiipiit of an LTIsystem is tlie bairie as the.orrelation pTiJ between the output and input of an LTI( T> syhtmi with impulse response h(tj? Evident 15.irne therefore has conjugate symmetry.9.106) Clearly.108) The autocorrelation of a complex fiinction of t.ion of time f ( t ) 9 S f ( 4 = P f f ( 4. to 9 by z is eqnimlent to a shift of g with respect t o f b. it follows from (9. m d the impulse rehponse.9. ' p f s ( r ) can only be expressed by ygf and not. 'The crosscorrela. because a s h i f t of f with respect. Exchanging f arid y in (9. by pfy itself.02) and substitxtirig variables yields P & ) = 'p.107) imtf for a real funct. (9.f o j * A J b . convolution of the auttrcorrelation of the input signal. with it delay circuit such Illat h(t)= (r(t.E* (#.v .he crosscorrelation fiinction is closely related to iiit. Mie will consider ail example. where the orosscor rclation is a hhiftcd version of the autocorielation Pyr(Z) ~ ~ ( r .2 Symmetry The question of the symmetry of t. . For f ( t ) = y(t).ion. the autocorrelation of a reel fiiriction of time is iln even furict.f(.9.2.
convolution relation (9. Evaluating t ~ w iii"icgra1 .110) \Frith the coiivolutiomi tkieoierri it f ~ l l f rt o ~ (!I 104) tliat m ~ YSY ( 4 0  f(4 0 'k !$*(d 0 %O") or c:ontisely.littion is not cornniut a The Fourier Transform of the ~ r o s s .112) Pm 1lie Fourigi 1ransfi. = wf?/Y(4> (9.riri of t l auto('orielation fmction this expression c m bc w simplified to ((3.230 9.14 we try to calculate thc autocorrelittion ol . the (. close rclalionship with cxdiangiag the t i r w fimctions Mtrin (9. we cif11 read from the general whernr ((3. that as with (9 96).r(f) = (9.109) in coiitrast to convolution.L +id4 f: P c / f ( 4 .rosscorrc. This meaiis tivr 9.4 (9.106) it c m br imrnetliately read that in gPiit1ra.61) (9.922.99). T h e Fourier ?i.arwrsforixi Cornrnut ivity While investigating the syninietry propmtics of the crosscorrrlalion fimct icm we Fonnd a.~ ~ ~ ~ e l ~ t i o ~ i The Fouricr transform of the crosscorr('liLtion tiniction (wi be most elegantly determined froni tlit.104).11 1 ) (9.1 13) si (i) Geiieralisiiig Exaniple 9. %'e need for this a special syminetry of the Fourier transform.
iiifinitely.115) we obtairi with clcterrnined in (9.115) I p. .101).18 we made good iise of this proper 1y to tl iirie how long we needed to nieabure in the thedoniain. There are vxious possiblities to defiiie the length of a bigrial and iCs bantlwitlth for general signal. 1 The corrcsponclirig sp ra are all. the spectrum of a rectangle fimvt iori does seein to become widn. We can read from (9. Despite this. when llie impulse becomes shorter. T1: ilarity throreiii (9.lic spe>ctriirnit. The autocorrelation function p ( r) then iollows by inverse trarisforinaliorr . The width o the spectrnm is kiiovrin as (lit. liowever. and ~ l l bee that they all lead l o the same funrlamcntal priiiciple. l ~ . in a certain way. = 27r the saiiw value that we bad ct A t various tiriies we have seen that there is a close connection belwcAen the dura. TimeBandwidt h Product 15 231 rlgaiii so cwrnplicatcd. exactly siioulrl be understood by the tc\rms period and wiclth. ~ f u ~ 1jilt we must make it clem m7hat. of the signal z From (9. strekhed out.9. 1 1Figure 9. to get a clear rtsult e freyiieiicydorriai~i. F { p I l(I)} i 3 = Ix(.tt. Here.s 511 rixii.ion yz.r ( 0 ) is t>lie siinie its the energy E. tkie value ol the stitoeofrr!l.. W e are cspc~ially interested in t h e r~lationship between the period of a signal in time and the witltli of the corresporiciing spectriini. In Figlire 9.tion of a signal and the appearance oi i(. We will consider three of them.113) thai the Fourier transform of our autocoirelatioir function is drwidy kiiown froin the Fourier t raiisforrii of tlrr sifiinctioii.G9) 5 that a sfretchtd t i m e signal corresponds to a comp spectrum (arid vice a) for any time signal.clearly iiot.w)12 = 47r2r (9.10. however. We iiow wish to investigate this rtmricctioii rnore closely..This reciprocal roimection bc%uwn the length of tthe signal in time :mnd the width ofI.(z) = 2nsi To check tthe results we colisider the value at t = 0. limittd to thc rcctanglc functiori.7 it is cvident what is meant by thc duralion of a rectangle function.(0) = E .. that w e would likc to find an easier way.114) (9.
that has the same height> a i d area as X(.116) The area.232 9. . The Fourier Transform EqualArea Rectangle Thc f i r s t method for defining the duration of a signal and its bandwidth is shown in Figuie 9. Dl ariti height s(0) z(i)dt = D l ~ ( 0 ) x we directJy obtaiii the signal duration D L (9.jw) itsclf. tirrtler the signal x ( t ) cannot only be tlcterminecl by integration.23: Dilratiolr I h and bandwidth BI of it 1% itnctrrstanti the width N I of a spertrttm X ( p ) as equividcnt to the width rectangular spectriim. This is The sigml duration c m therefore also he expressed simply hy thc ratio t Figure 9. W e understand thc duration 111 of a signal c ( t ) as the ciuralioii of a rectangle fiiiiction that has the same area and height a5 3 ( t )itscli If the time zero is chosen so that it coincides with the inaxirmtrn value of r ( t ) then the lieight of the rcctanglc function is equal to ~ ( 0 )From the eqiialarea requirement for the signal x ( t ) arid r . the ) Fmlricr transform F { s ( t ) }= X ( j w ) has the sanw properties. Conscquently.23 for a signal ~ ( tthat is real ancl symmetrical. hilt also by reading the value of the spectruni at w = 0.
ancl it is then For the handwitlth B therefore.24 sliows examples fox a signal and i t s speclxum. They peririit made about a signal or its spec1.119) B is clefirted accordingly z lx'(JLo. Figure 9. so for their prod uc't 111B1 = 2 n . 1 B1 = 2T.120) 2 IS of thi? kind me common in filter tfrsign. TimeBandwidth Product 233 (9.x (0) X(0)' Diiration ll. TI1e procluct DIB1of duration a n d bandwitltli is ciillrd the fzmebandw~dtlr From (9. A fixed for value or lower limit is unknown in tliis t 5 q Illax / x ' ( ? d ) l B .rl The bmrlwidth 'dt 6 [ t o .rum.rly 3 have a reciprocal relationship. ~ i can also he expressed by the balue 01 llre ) signal .10. f o + 021 . even when t h e exact behaviour is uiikown.mclwidth that we have chosen here: eqiialarea rectangles foi all real and syrrrmetric. (9. A differ cnt possibilty for tletiniiig the ciuiat ion and bandwidth t i rameters.0. and banciwitllli 1 1 (*lrit.tltimc signals. the limcbandwidtlr prodiict DJBz a cc. (9.rt:iin value of q only dcpeii on thc form of the higrid1 z ( t ) .117) Y 2T Ic ( 0 ) 'I'he area under the spect.118) product.(t)l5 q rnaxl. From the similaiity theorem.rum X ( p .9.J 18) we can see Ihal: This is true fur the definitions of duratiou and h. (9. Ontsicle of t h P duration Dz.. I ( [ ) at f . t h e magnitude of llrc signal s ( t ) is always It (0 < y < I) its maximum valite: IJ.)/ 'd lull > 2.
234 9.17 Signals of the form n. is eoiisl a111 for any given q.ers 1)ecorne strict.s.5cd.(t)= elu”z are callctl Ga iiss s?npdse. arid for thc the timebandwidth product. D2R2 = 8 111g . anid {lie resultirrg timebandwidth product Dz 232 foi Gauss iinpulses is increa. . the valut.24: Example of 1olw cf~irrparameters Example 9. rheir transform pair is (see Appeiiclix Appendix B.tar.T w2 (9. The Fotirier ‘l’ransfortn Figure 9.121) Note that l n q is negalive. If q is made sinaller. the tolerance parmiet.3) euLf2 0 @f?g. Q z/.
.124) is thc first mornent of I. In this analogy. Thr bandwidth B+ii defined by (9. TimeBandwidth Product 235 SecondOrder Moments A third metliocl for defining the qigiial duratioii aird bantlwitllh ( using secondorclcr momcmts for I h e rriagriitiidc squarcd of the signal autl its spwtrum. JVc are assunring that 3c 00 (9. so that that it has energy E . (9. (9.127) .124) U the mean arid (9 123) as the standard deviation. = 1.126) 00 (9.r(f)l' It corre5poiids to the rentrr of g i a ~ z t y rneclianic~~ in if the time &>pendent vitriable i s zicneed as corresponding to pmitioii (t. 1 r(t)j2 can br i n t q x c t e d as the piobabtlitv diitribut ion iiitlepencleiit of t . If the reader preiers the prokxtbility tlrreory analogy.g thc distdiice along a I)cain).1 25) with (9.123) ii thcl momriit of meitia.123) where (9. and ls(t)12 as thmity.9. we normalise the signal ~ ( t ) we are considering. For siuiplification.10. This drfimtion is motivated by analogies with ranclom signals or rriecltanical sy st ems.
h qiiantiirii IIIC5. It is forrnally related L o the uncertainty relation from quaiiturri mecliztnics. aiid vice vcwa. good tirnc aiid frequency resoliition i s uiittl at thc same t i i n c b . iiics.1 Calculate the Fourier transforms of thc tollowing signals with the Fouriei integral. Note: b) Piuperties of the Laplarr traiisform arc in Cliapkr 4.The exact wine ol D3U%depends on the forin of (lie sigiial ~ ( t ) .t ) ) r ( t )== c/.oiirliisions. This is. givc also the Laplacc transforms with the regions of convergence. Shortwnig the diiration of t kit) higiial always inrreasos (lie haiitiwidth. It is tliercforr not possible to fiiitl a signal that has any deqired short duration and at lhe same t h e ally dcsired small bandwicltli.. as long B S it coriwrgcs.1 transmission a i d spectritl aiialy ( w e Ext~inplr 9.ciprocal. 1)uration aiid bandwidth of a signal are rr. Gausi impulre has it particularly g I width proctiirt U 3 f33 i is t. mininiurn The is found for a Gauss irnpiilse [ l Becniixc of forxnttl analogies wit. this eqiiation i s i11s0 called the unrer As .4 . arid Gaii windows clre widely eniplowcl. Exercise 9.11). demand of a digital traiisrnisdion syst eni. c) IMes for cnlculations with the delta irnpulsr are in Chaptrr 8. a) z(t) = z(t) elWof c) n(t) = &(4t) d) Z(t) = s ( .iriplo)yed whcncwr it IS irnpoutaril to pack i f s m i d i riicrgy as possible f ilia11 frcqucncy b a d ovei a sm:tll amount of time.3. From t h various definitions of duration and hatlitlwidtli and the results obtained we can cliaw ~ O I W i m p t ant c. This statement i s very irnyortitrit for signa. for examplt. In shortlime spectral analysis. a.JlJl . For coriipa~ isoi.
9. ( 3 ) Sketch z ( t ) for this choice of Exercise 9. = a) Deterininc b) Calculate WO so that the mros of z(t) lie at t = ri .5 h 3 t :x:(t) si(w(1t). Exercise 9 . IIW Note: SCPSection 9.4n.1 can thc Laplace transform be used to calciilitte the E’oiiricr transform (see equation (9. whose Fourier integzals do not have conveigence.4 Evaluate 3 { (I} using the Fourier integral (9.2: a) lX(. You will find some already lciiowii trmsforin pairs arid some properties of tlic Fourier transform useful. using gencralised functions. E Z \ CO}.9))? Justify your aiSwer. IJ~. n / W m> x ( t )$1. Exercise 9.4.6 Rvaluate the Pourier traiisforrri or :r(t)= si(lOn(f + T ) )xud sketch for 7’ = 0. Exercise 9.1.11.1).4 and a suitable substitution Exercise 9. T Find the fimctions of limc that correspond to tfic followirig Fourier transforms: 5jw 5 ( 2 ~ ) 2jw ~ X*(jLd) = ~ + + + 17 .3 Give the Fourier transforms of the signals in Excrcise 9. Exercises 237 Exercise 9 2 For tvhickl parts of Exwcise 9.p~)l arg(X(jw) 1 eid 1)) R r { X ( j w ) ) ancl I m { X ( j w ) ) Note : tlisplacerucnt rule.
~ ( ttC. With amplitude moctulatioil.sscrl t) with X ( ~ U I )Use the symmetry scheme (9.4.fil).4 . Note: the tliml transform pairs can be found in Scctions 9.10 ) The hignal ~ ( thas spectrum X ( j d ) .r*(i).9 YOLIR I P given the transhrni pair ~ ( t ) o .238 9.2. Civc Y ( j w ) dependerit on XCp) and sketlch it.ttc Y ( g w ) from Exercise 910 witJr F{sinwrf} and thc rrniltiplication t h e e ~erii.he dualit! principle to help calculate a) i r ( n b ( t ) b) + $1 J%&} c) F(sign(t)) .12 Use i. 1)) yb(f) = ~ " ( ancl c) . Exercise 9. The Fourier Transform Exercise 9.2.* X ( j ~ .11 Calcu1.3 and 9. 9.y( ( f ) = . ~ ( t = ) ( z i t ) + n c ) \in(+t). '? Exercise 9. Mliat axe the Fouricr ) ) transhms of a) ya(t) = c(f]. Exercise 9. X(jU.) 4 Note: h a l y s e siri(dTt) with exponential oscillation5 arid iise the modiilntion the orem. expri.
9. cliffermi iatiori theorem ol the Fourier transform. I I.15 Detexriiirir t hc Fourier transiorin of thc triangular funclion shown a ) by using t hc. Exercises 23 9 xercise 9. . b) by niiilt iplyirrg tli pcctral functiorrs of suitatdy chosen r vvliose corivolutiori give5 the tlimgular ftmction.
This Page Intentionally Left Blank .
all poles must lie left of the iniagiirary axis of the splane The imagirtary axis s = IW is part o i the irgiori of convergelice of C { h ( t ) ) . that we can also write here 8s Ff(h(t)) .We have already poiiitd to tlic close connection between the L.9). They would always have logaritliriric gxq)li paper ready for this twk. It is mostly usrtl e systen~sso that we c m refer t o t h sydrrri function instcad of the Lapl~~cc transform of the irnpulse W~POIISP. we had iiitroduced the Fourirr transform wiLh ?w and not w .so ~ h Eiturier aiid Lapiace tmmforms can be r in (9 0 ) . Bode plots arc’ used as a fast way of firitliiig the approximdp ireqiicwj responw from the pciles and zetos of thc system function. g ~ ri arv.y response Instead t lie Four ic:i transform. At one time. &I”) with the following properties = lH($&I)/.1) As stable systcms only have decaying oscillatioirs.a ~narvellous way of plots faster and more accur dcvelopirtg an irituitivr iintlc~rstartdingol. aiid llie freyt1errc.aplacc traiistoim. = w4f)}l.=. at tlric s t a r t of Chapter 9.q)lm>transform m c i tbr Fouricr transforiii. Bode plots werr t hc erigiritw’s stanclartl riiei hotf for representing a freyiieiicy resportse in graphical furrii. as computwi can clraw frequency respoiw Boctc. this is 110 l o i i .. Any way.v 1 Tli~ fiinciaineutul coriiwct rciporise i5 given hy the relationship (9. thc magnitude and phase of the Foinier trnnsforin I from the polezero plot of i t s 1.HCPJJ) H(s)/. h o ~ e v e z .=lu (10. so between the Crecpmiry response U(’J) = F { h ( l ) )aiid the values of thc system fturction H ( s ) = G ( h ( t ) } for s = j w . in 1 h i s chapter wc iritwoduce a classical method to d rniinc. i h t s are. t J ” 1 # i H L I ” ) l . w e do riot haw to make any changes in the notation. in C h p t e r 9. Bod~ plots represent the frcquency rwpitnse c2s split irilo magnitude arid phase H ( p ) = I H ( ? U ) /. how the lo( ation of thcl poles ltritl mros rn functioii affect the fr equcncy response. arid it IS therefore iniportaiit ior developing and analysing 11 €uriction and the i‘rrqueiic. but of course.
Rrid 20 loglo 1 H ( J U J ) troni Table 10.ation. The pliasc ~ ( J w ' ) a r g ( f I ( p ) } is plotted linearly against tlre logarithm of L the frequency.SOnJ ( 10.242 10. P . Uocle Plots 1. Thc magiiitiide frtqutncy response is shown 1.a) folltrws log IN(gw)1 P(. cleu that tlie niagnit ude frequency iesporisc car1 be approximated by two asymptotes: for w < 10 the magnitucle < frequency rcsponsc is approximately coiistaiit and for w >> 10.h we wish t o tieterruine tlic contribution of a real polc at s = 10 with the trecpency response (10. In tlic following s ions. The linear represmtation uhows rieit'rirr thc behaviour at low lreyut:ncies (0 < ~j < 10).p)I are marked in decibels (dB).ttier when drawn. and then put the partial results to&?tlleK as corqdet. ivi es an s We will start with ail txaniple. in Figure 10. iii wl. ~ L I J ) \ . \ H ( . decays by 20 . 2 . appioximlte asymptotes fo1 tlic plots can be clraxm very quickly siwe f r m ~ n ( 7 w . in linear arid in tLnhlr logar ithirric reyrrseiit. 111 coiitrast. The logarithm of the niagriitiidc [ N @ J ) is plotted against tlic logarithm of ~ the frequeiicy J . we write 201og.?W) = 2 log jJ" : TIl n 2  1 n log 13.S1"' I + log K (10. arid the individiial magnitude.ti Bode plot3s.. For traide:r functions in rational furni. of clarity. we will first hliow t h c rcpreseiitatioii of siiigle poles m d zeios.1.1ic.%n)  z arg(*/" : 71  The logarithm ot the magnitude arid phase call tliiis bc separately determined for each zero and pole. it. nor tlie bchavioiir at high lrcquencies (w >> 10) particularly clearly. A logaritliinic frequency axis makes i t easier to get an overview of multiple decades. tlie rloublr logarit hinit rcpresentdtion makes it. i.1' .. can br added togc.1. Likvwisc. the ordillales foi log IH(.e.3) ' = a l g ( . for reason:.4) 13v suhstituting different values for w we obtain the value for IS(p) arg(H(p)} 1..
where both asymptdes n the inagiiitucle of the pole at s = 10 c the 1 m Ir.iii~i~ci. form 201og.w 0 5.20.2 fox a s y s t w i i with a real pole at s .t ion i .y corrcspoiid to the vali~eII = 10 f r o m ( 1 0 .71'  20d I3 .L_" * 0) 100 200 300 400 500 101 100 101 102 103 Figure 10. The rruinerical values of the rriagiiitucie Gtqueuc.0 the frequency Pespoiise is purPly real ancl has the value E i ( 0 ) = = $E.Y 1* 20. N ( 0 ) = 20 x (1) = 20 (11). at and abovc the cutoff he( p r i c y is sumniarisetl in Table 10. The exact i d the nsyinptotic behaviour below. Thc fiec1wwy w = i 0. from these vcilclri arc' oht.10. At w T..1: Linear and doublr l o g a r i t h i c represent ation of t he frcyirencv responw dB/tlecade. the phase ari& is c o r n for imy real poles.1: Xfagnitrdc.o .gup71cy tilc hystcm. . I7dt3 logarithmic I0I \ 1 40 50 0 ___ .23 Table 10.n. or in logarithniic. 4 ) . phase and dampirrg of H (jw) for vauions fieqixerici~:. The syst em hit.2. Contributhii of Individid Poles aiicl Zeros 2.. The phase angle is Or' At rising tueywncies IIO significant variaticm:.04dB 1I. : 1 N ( s ) = _____ s+cr' n.
.jw)I arid log U ) is oblainrtl.1(~.1 = 6" .214 10. dainping ancl phase Frcyueitry H(yw) 1 1 10 Damping Pharc 0" M U) =0 = n 20 dB w < 0.2: Table of values of 1 f ( j w ) .1tr w < a ou w = (1 w>a (*I 1 1 x=a 10 1 Ja 1 a 1 N N 20 dH 23 dB 20 dB/'decarle M 45' 3d > 1OCX 90" to about LL. Bodc Plots 'l'dde 10. liirear relationship between log lSS(.an = . = O.1a At this frequency the phase has the value p(.jO.ailcl is given by i i t j to u ) ~so that.arctaii0. 0 O. Incrrasing the .70.0 at the cntoff frequency w = a. hi the doiiblc logaritllrnic representation.ln The magnitude of the fr~quency response first varies significantly from the value at w =. because of log IH(ju)l = .la) = = ~ arcl.log LJ. it is (hen decaying reciprocal to w.) arg H(.
Its exact path rim:. Between LS = 0. The phase has the constant valiie (1" for w > Ion. arid afterwards it.ln. The advantage of this representation is t h a t the cxact path CRII b t ~ skrtched from the asymptotes arid the known deviaiices at O. innderncatli the asymptotes and its maximum deviation from tlicrii is 3 dB.arrtari __ = .la a d w = 100. ~ . The phase reaches the valiie p ( jIOU) = arg N ( j l 0 a ) = . It is obvious that the shape of tlie magnitude frequency itsponse arid thc phaw c m bcl ctesrribed itpproximately hy a few rides.2. the plias~ decreases linearly with the logaritlirn of frequency. at the cutoff frequemy of w o. magnitude 20dB 23dB phase w 101 100 10 101 100 101 102 103 Figure 10. The greatest variat ion between tlie exact path arid thc asymptotic appioximation is about 6" arid it occurh a t w = 0 lcr arid w = 1Oa.10. The phase climbs between w = 0 .arctan 10 = 84" U 100 = IOU. Tire magnitude frequency response is sdd to dccay lineady by 20 dH /decade. Contribution of Individual Poles and Zeros 245 frcqucncy by a factor of 10 (oiic tlecaclc. 1 a i d w = I O U from 0" to +gou. n arid IOU. at iu' a The magnitude frequency response is coastant for freclueiicies below LIie cutoff frequency w = U . falls at 20 dB/tleradc.2 shoxs (lie asymptotic approximation (by solid lines) and the exact path (dadled lines) for U = 10. Figitre 10. At the cut off frryiicnry the ph' has the \ d u e 45' exactly.For w + ocl it hecoines W.2: Bode plot for H ( s ) = ~ s + 10 1 FOIa real zero.) then leads to a reduction of log III(jw)l by 20 dB. the s & mrules applv with a few deviations T h e magnitude frequency response climbs above the cutoff frequmcy at 20 dI3/dtwde.
4 ahovr shows tlie ' logarithmic represeritatiori of the rnagnitudc freqiieiicy response from the rule:. in thc last section. (10.o f f freqrrericies are w = 1000 for the zeros. tlmt is for a transfer function of the form N(s) = s i. aiid w = 10 for both poles. pole by flappiug tlie tirm4ng along the horizontal axis. 'rhe value of the constant asymptote can he most easily obtained for w . Rode Plots Tlie plot for a lea1 zero can be obtained Gom that for a.10. In Figurc~10. The roritribut ion of Ihc x e ~ o can he giveii irxiinediat4y by t i m i n g iipsidc down the contributioii of thc pole above.246 10.10.he linearly represented phases togcther. for syhtcms with riiultiple poles and zeros has already been done. tlie phase plots ale shown. Because ot tlie greatrr cutoff frequtxncy of w = 1000 it is vhif'ttcd two tlecatlrs to tlie right.2).. again by iiappiiig of tlie pole as above.3 shows the correspoiidiiig r i i i i b for a zero at s = U = . The contributiom of' tphepoles cmrespontl exactly 1 o the syrtrm fiirictiori (10.3: Ikide plot for M ( s ) = s +10 After the detailed examination of the coritribiitiori of individual real poles a d zeros in the last section most of the work required to clraw a Bode 1 h 1 . ____. Thi. and by shiftiiig two tlrcatlen to the riglit. Figure 10. 1000 = 60 d13. magnitude phase (13 10' 100 10 100 101 102 Figiiro 10. Tlic c i i t .0 to 20 log.4 below. We will explain the procedure for the system function I](. The phase of thr zero caii be obtained.3) can be written as the s l i m of t'nc contrihutioris of the iiidiviilual poles anid zeros. we obtain the cornpl . we only riccd to deterniine the the poles a r i d zcros arid piit them together to foim w cmq)Ietc pictiirc.) = s (s i1 0 ) 2 + 1000 = (s + 3000). phases of thv polw h t w dready kriowri.4). As systems wil h midtiple real poles and zeros (10. If wc add cacli of the logarithmically represerketl r~1i~gilit1~&cqiicncy fi sporises arid t. 1 1 SCl0 s+10' Thc double pole will lse tieated as two siiriple poles. Figure 10. that we have already dcdt with in detail.
4: Piititdin(: together a Bode plot for multiple poles atid zeros.5: Complete Botlr plot for H ( s ) 7 h 1~ (s +10)L 1000 . from the cut.1 x 10 = 1 arid 10 x 10 = 100 by 3 X 90" = 1800 from 0" to . In (he sainc m m y l. Biginr 10. AS shown below in Figiiie J 0.1 8 0 O .off frequency of U: = 10.3. Bode Plots for XiulLiple Poles arid Zeros 217 20 log IH(jo>l [dBl 20 0 20 40 60 Figure 10.5.10. Bode plot.he phase decreases between 0. The double polc as the suin of two simple poles leads to a rlecay of thr: magnitude freqiiency Iespome by 2 x (2OdB/dccadc) = 40 tlU/tlecade.
I ~ a ~ Frequency n ~ ~ ~ ~ e 1. 4. Dcterrnirie the locat. Extension t o complex pairs of poles and zeros i s covered in furtfiel sections. As Rode plots are vcry ra%yto draw under these conditions. Lilbel the vertical axis: calculate lH(jw)l in a region where t. Straight line to the next cutoff frequency. thcn contiriite with step 4 until all cutoff freylteacies have been finished with. increase by 20 dB/decade for each zero.ui.erminirig Bode plots with some simple rules for the magnitude frequency respome and the phase. Start at srriall w: :$) no pole and no zero at c) zero at s =0 s =O + i gradient 0 gradient 20 dB/decade + gradient. The* liniitalion to real poles and zeros has historical reasons. 2. they are the most tviclcly used. They apply to poles and zeros on the real axis of the left half of the splane. so only tlic contribution right of the c. . W e can see from our previ011s discussion (see Table 10.off frequency appears.2) that a factor of 100 is sufficierit. Draw axes and plot cutoff frequencies 3. howrver . that is the gradient of &20 tlU/decade for the magnitude freqiicucy response and &SOo for the phase. ules for lot S We will now sumniarise the discussion about det. Bode Plots The increase of the phase of 90" between 0.re a pole or a zero lies at s = 0. 7. Decrease by 20 dB/dccacte for each pole. difference.lie Bode plot is flat..248 10.ioii and order of poles arid zeros. the cutoff frcquency is iiot visible at ui = 0. +20 tlB/dcmde For rnultiple poles or zeros use multiple gradient b. 5 . The following rules also include the case whe. The d separating them should be great enough so that they do not mutually iiiflucncc each other in the region of tlio cutoff freyuericies . Round off cutoffs hy i 3 dB (or multiplw of f3 dB for niultiylo poles or Z"l"0S). 6. In the logarithmic form o f Bode plot.1 x 1000 = 100 and 10 x 1000 = 10000 leads to a total phase of 90" for ui > 10000.
The Bode plot with thc correspondirig cut off Frequency can be dxawn as before. I lriiritis sign before H ( p ) is a phase offset o 180". Smooth the phrase sketch. BiJttr in niirrd t>hat . .ie cutoff freqirerrcy. tilt* ciitoff frequency deperilds alrtiost cxclusiwly on 1lie itnagiiir~rypart.40 dB/dcc1l. phase angles. because tlicvi the rvtoff frequency is almost exclusively clefiiieti by t. Complex Pairs of Polcs ancl %ems 24 9 1. 5. 1 the cutoff ~ Treyrienq to 1Ox t. Cloritiiiuc with step 3 tnitil 011 cntoff frcqueiicies have been dealt with. It. 3 Start at srnidl u l : a) no polc aiid n o 7ero at s = (I 1.de. and zeros there are 110 simple rtiles as with real poles a i d zeros. s = 0 c) zero at s = 0 + + phase 0 phase 90" ph. The equivalent is also true for pairs of ZeTos. the r ~ apart o a pole is sinall comparcd t o the iriiagirmry l f part. Dcterniirie the location and orrlcr of polcs ant1 zeros. E:wh pole siihtracts OO". a c m q h x pair of poles can be qq)Ioxiniated izs a double pole on the rcal axis.5. A sufficiently l x g e irnaginaiy part iirclatioii to t.1x the c*uli)fffrqiiency and 1Ox the cutoB freqiieiicy (or the c. carlt zero adds 90" in a region from 0 . rhc equivalent is also triie for pairs of zeros.orrcsponding multiples for niultiple polc.tse f90" h/hdtiplc poles or zer. hccausc the cutoff freque1ic. The belicLvioiw at the cutoff frequency depends o r 1 the relationship between the r e d part and the imaginary part.hc i eid part.l. on the other harid. For complex pairs of pole:. I x the cutoff frequency. Straight liuc to 0. Rouiid off about 6" at 0.he real part ('I(" i resonancc thaL becomrs oiiger as the pole iiears t h r irnaginarry axis. If thc imaginary part is siriall conrparwl to the real part. G.) pole at.s or zcros). As we are dealing with two poles.10. 2 . Diaw axes and plot the cutoff frcqiieixies. this amounts to a gradient abovc thc cutoff frequrrrry of .os h a w multiplt. f 4.y d~pends both the real and imaginary on part of the pole or zero.
04 s = 0. T h c torrri of the resonant peak dep(3ntls cm the distairce of the pole from the irnagiriary axis. Bode Plots Example 10.__.ui. je naher Pol an imaginarer Achse lie@  40 dB J Dekade \ \ I 1 10 100 Figure 10 6 Rode plot for a complex pair ot poles The cutoff frcyuenry is in this rase mostly deterrniried 11) t hr imaginary part and is located at w = 1.6.o .2 i'7 .To thc right of the (atitoff frequency there iy a lirwcrr &my of 40 ctE/dccadc 1 20lOg1()lH(qLct)/= 20log.ion H(s)= with poles a t s2 + 0. To its left tlir riiagnitucle fiequcncr response i s approxinzately coiirtant at 20 loglo 1. by the .. and the exact foim of its path xicar the rutoff frequency must again be d~t~lKill~led distance of tlw polce from the imaginary axis.04 M 0 dE. '\ I \ Resonanz urnso starker. 'rlie exact behaviour is depicted iir Figiw 10.250 10.= 30log. ~  .~. The phaw clecreases in the region of the cutoff frequencv from O* tzo180". ~ ~ .48 1 + 1.1 As it11 exariiple we will consider the inagriitude frequency response aid thcl plrase for the system finict.
Pole.7 sliows the sil uation €or x pair of poles and a n arbitrary point 011 the iutaginary axis. c1  ' U Figure 10.pro tli<igram €or thr given tmnstex Eirncl iori To estimate ~ J I mijgnitude G e q u t w y response we begin by consitlrring the P value of the coritribution of tlic system furiction at.x pole pair Example I As ail exaiuple we will exaniiiie the systenr function wit11 the polezero diagram from Figure 10. Cornplex Pairs of Poles arid Zeros 251 If cither the real or imaginary part may riot he neglected in favour of the other. We can in fact look back at thc equations (10. The distiGliw to the poles aiid tlie col responding angle c 2 1 easily be read from the plot.8 Figure 10. I. = 0. tik. Figine 10.5.liere is 110 simple method of finding the ciitoK frequency and drawing thc Bode plot.oiii the polezero .3) arid det cmiine the tfistancc of the poles and zeros.8. and the corresponding xrig1r.s grtipliically for each point of iiiteiesc on t h ( x imaginary axis.7: Polext*rodiagram for a complc.10..
we c m est. changes sharply around w = 2.90" = 2 0 " . if the 7 .e the phase a t the following points with the polezero diagram. the magnitude frequency response lar > 2 fah at > 20 dU/clrcnde.10.1. the lollowirrg apprtrxiniatte v. (10.zhies are yielded. In order to apyroximatc tlw path more accurateiy.2 . As the rrunher of poles is by one grratcr than the nunlher of zeros.30" + 90' .9. I/'lirtlncxr. as shown in Figure 10.5) For the pliase plot wc need the phase at J = 0 and recognise from the polezero diagram that diie to the pole s = + 2 j the phrtsc.90" == 90" p(j2) =c: 2 x 80" . To confirni this w~ determincl the rringnitudc at d = 10. we ~ a l c i i l a l ~ tdur fm L the 1 8s wrll. Considering this. Bode Plots plot W P recognise that at.irrtat. U = 2 there i h a resoilant point. We obtain 4 t7(. iTfe thus calcuiak the riiagnituutlr for this point from H ( s ) .akes a value of 90".50" .?0) = 0 ~('1) M 2 x 60" . The cli. As is evident in Figure 10.0" . the phase of H ( s ) for w >> 2 t.6) .lmces in equation (10 3 ) are calciilatcd from the polezero plot.c 4 ( I U 9 *t w=0 ( I w1 ul=2 w = L0 Fipirr 10 9: Estimating the lengths €or various w distances of the zeros are multiplicd and divided by thc distmce of t h c p o l ~ s : (10.252 IQ.
. our estimates of the phase as clisplayed with the exact values for corriparisori in Figure 10.10.12. Complex Pairs of' Poles and Zeros 253 + 0" w=0 w=1 Figurr 10.10: Estinietioxi of the phase for various w In Figure 10. Figure 10.. rhc estimates are rriarked as crosses.5. oo . Likewise.12: Exact phase arid estimates for various w . _ /' i Figure 10.11 oui estimates are displayed with the exact frrecjriency rrspoiise l H ( j w ) l .11: Exttct ntegnitude and estimates for various w arg { HGCO)} 90" x.
knowri as the Qfactor is obtained by relating Aw to tlw rwonant freqixeiicy W O . + + ~ d ‘I‘he resonant frequency is w‘o l’igure 10.tioa with poles S=Qkj(3> n polynomial t.akes the form (S ap 0: i s shown in Figure 10.he stratrd coiidition < /j the < deiioiriiriator t(.13. A dirrierisinriless yimniity. 5+ (2 p 2 ) M 2 + z a .13. the niagriitride rcsporise can also be cspressecl directly by tlie tlenoininntor coefficients.1 arid B = 1.13: Polezero diagram with pulcs very neai the imaginary ctxia The rtiagnil tide respoiisci plo d in linear h i m is shown iiFigwc 10.j$j(s + Q 5 j p ) = 2 + 2 ~ 2 .Y . This kind of frequency response occurs for all oxillating systems with low danipiiig and is called a resonance riiive A nwiwrc for the resoiiitiic*e is the width of the resonan(c cinve 3 dB below the peak.14 for cy = i 0. 1 (:an b~ cmifirrncd that 1 the eiid poirits ocmr at w z /3’ z t n arid Clie width is therciorc Aw x 2 0 . This cizn also be illustrated easily as i i Figure 10.For secondordcr sysi pins Kit11 a complex pair of polcs vcry close to the imagitiary axis. The Qfactor i h giveii by . Under t. T h c polezero diagram for system func. if a trimgle is tlrawn bctween i two poirrts i3 i CL on the imaginary axis and the pole.
2 How much amplification in dU does a system have.14: Resonnnrr wrie of the system xereise 10.10. First calculak the. . if the oiltput pomw i s :+) 64 tinics OP 1)) 2 times ~ h tinput power (thci input aird out piit resisluirccs arc cquual)? .6. relationship between input srrd output voltage.1 For the following sysys~cm: xercise 10. Excrcjscs 25 5 E’igunr 10.
500 a) Give H ( s ) arid draw the Bode plot.I 100' of H ( s ) in dB for h e fretpencies w = 10k . b) By what perwritage does the exact amplitude differ froin the one in the> Bodc plot at w = 100 ancl = LOOO? .7 a) nr av. use the ' frqiiency w =3 sz 100 For labelliiig thc axis. I. .l 1 10 100 14 Exercise 10. aniplitiide and phasc plots with appropriately labelled axes. .4 Draw the amplitude and phase diagiains for tlie Bode plot of H ( s ) = d si1 ~~ To find tlie appropriate scale For &lie arnplit iidP axis. i.01.e.7 the aniplii ude diagmni of the Bode plot for H ( s ) = (9 + I0)L . The cornples valiies of H ( s ) valiies should be given in Lhs form ci.3 1000 For H(a) = _______ calculate the the arnplifieation of H ( s ) in clB and the phase s . 5.0 50 1   999 e '~(t). 10 and 10'. Exercise 10. Bode Plots Exercise 10. calculate the amplification at = 10.1.5 s(s i 100) '  Draw a polezero arid phils? plot from the following amplitude plot. . k = 0. b) Give tlie output signal y(t) of the systeiri for the input signal x ( t ) = cos(wof) at ttie nornla~isedfrequencies W O = 0. A syslenr has the impulse xespoiise h ( t ) = b(t) .256 10. 20 log 1HI 0. j b . Exercise 10. + ~ Exercise 10..
be ~ ( t= ~ ( f ) .Detcrrriing the valuc of ) e) J A tlrc input sigrial of the sys the output signal y ( f ) for t i 00 from ail aiiiplitiide plot.1 101 1 2 llOIOs 10000 + + + a) Draw the polezero plot for this system.z percentage.10. Exercise 10. xcrcise ~ ~ ~ 1 1 The gi%inof a system a} decreases rccipracally .s of the Uode plot of H(. What. Exercises 257 c ) H y how many dB is it rounded oiE at the cxt.off frequency ul = 10'? Exactly 1ro)w much airiplification i s yielded at U: = 10:' Give the cliffercIit*ef r o m llre rvunded value in t h c Bode plot as .1s 0. 13) Draw tlrc. behaviour does N ( s ) sliow? f t t w l the iriaxiiiiiiiii gain aiid tlio freqiiency 01 frequenries where t he gain has fallcn t o a tenth ol its maxiinum.1.10 Mark the cutoff frequerrcies in the following polezero diagram and draw the ningiiitude oomponcnt of the Bode plot.6. rrregiiitudc ancl phase coinponmt.2 diKer? + 1 ) 2 and H 2 ( s ) S2  ~ S2 ( st 1)2' Wlicrc do the e followiiig transfer filwtion: H ( s ) = 10" S'$+ s2 4. Civc H ( s ) arid label the (HI axis so that IH(w = 2 x 10")l = 1. Draw the phase plots of H l ( s )  ~ (s aniplituclc plots of' H I a i d H.7w')with appropiiate labelling of' axes.
H j p } increases for ul << by 20 2 and clccrcnscs for W' 4 3 0 by 2 d13 0 ~ . d 3.tvcl signifies a donNote: Exercise 10.. ~ b) A firstordcr highpass filter has been connected to the systein. RI aplricnlly from it thc values of A a i d w'.258 1)) increases lineaily 10.12 Give the gain IHl of the system froiii Hxcrcise 10. trtinsfomiing filter with the following characteristics:  max. 4 0 g and 6 0 g in bliiig of frequency 2".5 at the frequcmcics d = 2 . Bode Plots c) incrtwes ciuadrttti d l y cl) iriLie&seswith order rz with the frequency. oct.15 The followiiig i s known about a system: Qfactor 7 50. = Exercise 10.gairi of 0. WO  riiaxirrium gain i s 26 d U at 10. A.. 2 0 $ $ . gain 14 dB . j Exercise 10. Giw gain Hirp( s ) and the corresponding amplitude coniponent.5 atJw = 10 The higlipass filtrr has cu(of€ Erequenry dSi arid fox U! w .14 Colisider a secondorder lon7pass filter with cutoff frequency J.ul = 8 and ul = 2000. gaiii of 20 dl3 2 10' itritl a 1)C a) Draw the arriplitiitle coniponcnt of the Bode plot with appropriate laltellivig of axes slid give the correspondiiig t r snsfcr funrtioii N L(8).13 Give tlip gradientrc. How iiiaiw correspond to this freqrienty rclsponse? Exercise 10. then draw the axnplih i c k comy)onmt of the desired systcm ancl cleteirnin.
6.10. Exercises 259 .
This Page Intentionally Left Blank .
g.7)is called tlie samplznq %rLte’IYJCll.. because the amplitude is rounded to a firiite word length.~ l ~o f ea coritinuotis sigiial by the duality prinriple of the Fourier ~ s tra nrform.1 The coxitiiiiious signals dealt with in previous chapters occur in natural and technical processes. The square brackets iii r[k]iidicatc that it i s a discrete signal. This is carried out with an analogtodigital converter (ADC) in two distinct stcps. hniques and the results obtained can be elegantly carried over lo deal with sampling. the appioxixnatioii is in mmy cases justifiable so we will foregu discussion of it. 64). the sampling. a bequenw of imnil~ers. i. a Sarriplirrg of a continuoustime signal ~ ( 2 at equidistant tiirir p o i r h sepa) rated by a time 7’ (time quantisation). = Storing the sequence of’ values ~ [ kin memory cells with a finite riimher of ] bites(amplitucle quantisntion).r[lc] z ( k T ) . 11)+”8 to 10’30’) and a correspondingly high number of bits (e. or time quantisation.hem to digital signals. ‘1’0 rc~ortla i d process these analog signals with a digital computer we need to coiivert (. The time T between two sequential samples r(k. . We are only concerned here with the first step. The first tool wc introduce for describing the sampling process is the impulse traiii. k E z. For computers with a sufficieritlj great storage range (e. . whose iise we will tleinorist e on a known problem: iepreseritiiig a contini i o i i s pc%riodic signal with a Fourier series.g. wherc tinic (or position) is a contimious vitriahle.e. The line spectrum of a pcriodic signal is closely related to the sample v i . Arnplitude qiimtisation is noiilinear.
. As each individual rxpoiwritial tcrrn tJI"' h.2) is corrcct btit iinpractical.Signals Wc under*slanddalta iinpulse tram as the b u r n of an irifinitr n u n t l ~ of shifted delta r irripolses. train The Fouricr transform of the dell a iiiipulse train ran be olttaiiied simply by t ramforming tiir individual inipulses with t hcl displaremrnt th r() r ~ i i . RA shomi in Figure 11. Sarnrtlinrr.(> cdlecl the *&asymbol' . In order to dcrivc. wliicl~has t nt. 1L &4n. Although the product X(JUJ) xsi(ui/2) .1: Delta impnlst.1. impulse traiii i s given by llie forrrrula thc X(t) = lU(t) = \I= 6(t .1) Figure 11.262 11. . B ~ r n i i s rof its sirni1:uity with oiic of the lctters in the cyrillic . . x (11. ihs a shoit h i u . Likc t h ( x delta impulse itself.~\period 27r. in nsc of distribution theory. a mole kic)ipfiil expression. we iiiiiltiply X ( .w t l Periodic.tlpl. it is alt. X ( p ) = X ( ~ ( W 27r)). syiiilool i i i ( t ) . it is a g c r i c ~ uliscd filiict ion. because of the iriiiirite sum. Wc will lw xising tht.iabct. 4'The form of tlre spectrrrrri S(J. llsing this sy~nhol. foini as t h impulsc train. fbr which t h c distance 27r corresponds exactly to tlir priiotl o f X ( p ) . obtain and (11./ I ) .. j w ) with a sifunction in t1w f~Pqnericydo~riatin.3) The funclion &(&/a) 1 i S zeros at w = fh.J) troni (I 1.
j(w 2 ~ ) deteiiiiiriecl with (11 2) coriit~s ) into use.2. (11. . (11.') is tb fore cqual to a delta impulse with meeight 2n. By coiriparirig (11.3) i~ = Wc c*oiiitlnow tI(+mnirie X(7w) from (11.hiit siiice si(w/'L) i h 7ero at.3) aiid (11.8) . As rect(t) lias width 1 wliidi rorrrsporrds exactly to t h e interval betwwn the irripulscs in i i l (tj. a very sirnple fuirction.atcd.periodic. . +4n.11.. tIie points 1 2 n . and the viilur of the product is liltewise zero we do 11oi know what the v d i i ~ X(guj would be.is) it can be confirmed that + X ( j w ) = 2.I/ (3.a). continuation yicMF Here we liave used the scaling property (8. ~ ' ( J w i:. Inverse Fourier traiIsforrriution with aiid convolution yields y ( t ) = U ( t )crcct(t) = I .3).4) 7 hc timc fuatction initially looks morr complicated than it 5 spectrum. . tlir pcriodic natuie of of X ( 3 u = Y(.a).19) of the delta irnpiilsc (11. however. For these points. tlie coiivoliitioii U ( t ) * recl([) siiiiply yields the constant 1 Figirre 11.6) For (11.1. and we will cxplain lliis 1)y ) taking a cletonr through the timedomain..oiriplic. Delta Itripulse rYraiii itrld Periodic Functions 263 lookn c. Iiowever.r6(w) for  n<w 5T . Y ( p ) 'Lnb(b2) (13.
e Founer transform of a delta m p u b c : tram tn the teme d o m m i s a delta zmpulse trim vn the fr~qzicncWdoma2rr: (11. First we cwiisider a periodic.arid frequencydomain. We will 11ow use the delta impulse train to derive these. lines is given by the periods in the timc~tloinain The weighting or individual lines can bc clctcrrnined using Fourirr scries. 10 . impulse train (see Figlire I 1 31.we no gaps between repeti tions of t h function 1 ' (1). time signal s ( t ) with peiiod T .here .where the scaling property (8. repr nting periodic sigiials in the time. It can be represented as convolutioii of a function T o ( t ) with a drlta. To describe sampling procctlures with an arbitrary sampling interval T we use the delta impulse train (11. in particular. so bcforc we t.10) We also dcacribe these with the shasymbol from (11. Its use.niplc)y i t to cled with sampling continuous fiinctions we will try it out on sonie classical problems.1).264 11.13) T h e separation T of thc inipidsos is sct so that t. we can derive from (11. Sampling and Periodic Signals The asLonishingly simple result of this can also be expresscd as follows: ih. Periodic sigimls have a lirir spectrum where the di5t itnce bt>tween tlit.11) With thc siinilarity theorem.1). we defiiied an impulse train with an interval of 1 (11.19) is important (1 1. (11.12) between clclta impulse tiains in the timedomain arid frequencydomain is v t ~ y clegarii and can greatly shorten otherwise cornpiicated calculation^.9) the general relation (11.9) In order to keep thc derivation as simple as possible.12) urier ~ a n ~ f o r ~eriodic Signals e d 7'hc connection (11. howevri rcquires some practice.
3: Reprehelitation of st periodic signal z ( t )as the convolution of n period ~ o ( t ) wilh a delta iriipulse train For the Fourier I. The weights c*orrespondto the $ rcspective values of thc spectrum X*(&J). .rtirrr .4 shows this relathaship bctweeri the irniplse train arid khc c m i tiiiuous spec1..we obtain with (11. 'i'he relationship Betwcwri the prriodic time signal arid the Figure 11.14). = * t t Figiirc 11.radorm X ( J W )of tlir periodic signal z ( t ) . Figure 11. the integral of the irivrrse Fouriri transform .12) arid the mnvolirtion tlicorcni The Foiirier trarlsform X ( ' w ) is therefore as we alxeady know . Using the selective property.?U) of one perioct. Delta Impnlse 'I'rain aiid Periodic Functions 265 *. (.a hie spectrum iirc of w : between t h r lines.multiplied with 2n.4: The F'ourier transform of a periodic function is EL line spectrum intlividual linccl in the spectrimi is obtained as the invare Fouricr transkxm of (1 1.11. 'l'hc bcight of' the arrow symbolises the U weiglit of tphcimpulse.2.
on sampling functions o f time. .riotlic signal ([he inpiit signal) with a n apeiiodic signal (t.16) Equatirig this expression with the general lorm of the Fourier series 01 a periodic fiirictiori gives (11. W e will ref'e:er h c k 10 it whm WP de:d with sampling. 1 are interested % iii the relationship between t h t ~ Fouriei coefficients of the input and output signals. poiiding convolution integral i h s clemibes t hr corivolriliori of iz pc. Sampling and Periodic Sigiials ix b u m becorrirs of t hc individual freqiicncy lines ( I 1. ii7put signal s ( t ) \.1).oeffic. Exaniple 11 1 As xri exariiplP TW consider a periodic.dso the values of Th the spectrimi x i O ( .it11thc Fouricr series (11.16). Firstly.19) x .)a) the periodic signal of z ( t ) is theiefoore a line spectrum wlme the weight of thc individual lines is given by the Fonricr coefficieriis of ~ ( t ) . the output signal will also be periodic. p ) wit11 period ~ ( t )multiples U: = J/'. ixi5teetl or finding sitinplc valws from a spwtrurn. WP will d r d with corivolrition of j)er iodic and nperiodic +gals.hta irnpulsc rcsponsc). This compact derivation sliows that iisiiig the impulse train ant1 the sh saves ciiiiibt)(arbonie r. t hc firntliznimtal at of frcqueiicy (see Figiirc 1 1.altulation.17) WitJi ( 1 1. hoivrwr .18) 1 Eoi the complcx Fourier c. T h e tmpliasis will tlieii lit. we obtain (11. e Fouriw coelEicienls w e .266 ~ 11.ient)sA .15) (1 1. ~ ~of a Periodic and an ~ ~ ~ 1 1 ctw be obtained in t 1 The output signal of ail LTIs the impulse respoiw (s toiivoliitioii of the iiipiit signal input signal is piiotlic. The syectrum S (.
Figure 11.t & ( t ) (11.oefficimts are thc product of the Foiirier cocfficicnts of the input sigriill and tlw ern freqitency rcyonsc at the frecpiencies of the individual spectal lincs Cuiivolution of any pw3odir signal with per iotl T that cwi he rcprcscntccl by its Foiiricr coefficients Ak (11. with ail aperiodic sigiid h(f) yields .2 1) The Fourier transform of the inpiit signid is a linc spectrum: 2nz4r.6(w.22) Thc output sigiial likewise has a lirie spectrim. Its Fouriei c.airied from the sy&m fiirictiori for s =jw: (11.17).5.27rk). Comparing (LI. product Y ( p )= N ( J u ) X ( J W ) .20) as shown in Figure 11.and on LTlsystcwi with irnpulue response h(f)= C . systern is obt. as for tht.I!I) and (11. k (11.5: C'onvolutioit of a periodic iirpiit signal with a n aprriodic ixrrpulsie rebporise The frequency response of a.17). ohtiii.in wt' I ( 11I23) The final invcrse trar~sfolnt gives the Fouric~ series of tlie oiitpnt signal. we rcwgnise that tlir period i s 2' = 1. .
2Bd 11. the product of both priotlic signels iindcr thc corivolution integral i5 lilcttwisc periodic. but in tact. determined by ro is unimportant.29) . all of the> f informittion about a periodic fiiiittioii is ernibctltlcd in : single period. f f ( ~ u i ) is its frequency response. Pcriodic corwolution correspoiids to normal (*oilvolutionof a periodic signal $1 ( t ) with a period of ~ ( t or. and (11 2 4 ) spcrities how the Fouricr coeficierits changed by the system.26) tincl (11. It seems likely that the rcsults o l t h e last sectiou can bc rxtmded to coiivolutioii of two periodic sigiials with the same period. it rriust bc t possiide Io define a modified convolution by only integrating over one period. p ) = F{ It(t)).27) irito (11. The convolution integral o t a u periodic* signitls thercForr rkws not converge. We are again interest ed in !he relahxiship between 1 he coefficients of the Foiiricr scrips representation (11.2G) k (1 1. If h(t) is the impulse response of a11 r.27) Y First of all w e put (1 1.ion is likeMiisp periodic. coiivolut. Ii 2s crlso cnllrnl cyclic convolution. kiowever. ) f 'The resiilt g ( f ) of pexiodit. a period o z l ( t ) with the periodic signal x L ( f ) . Sarnpliiiq and Periodic Signals whcre N ( .25) m t l reorder the terms %o+T (11.TiS)stenl.i l s long A S t lie iiitegratioii is only performed ovei one period of lmgth T . rll!i~exact location of the integration limits. Siiice.
:j . For t3hespectrum X(3w) oi: the hampled signal I(() we IW the multiplication theorem (9. Sampling 'I'herefore for the integral.32) 1 Ideal S a r r i ~ l i n ~ Now that we are comfortable with the spectra of periodic signals and the delta inipulse train. the wcights are syiribolisrd hy the lerigt.11.29) viclds k Me orthogonal.(t) . % . with the Fourici. we find the coxincction between the Fourier coefficients of y(t).atcr we will see that the con( for realisation can also be described by sirriplr cxteiisions ot the concept of ideal sampling.al principles particdarly siniply.28) ancl (11.31). a i d is there is a diiality her(. WP will use' first dertl with ided sarripling.6. 269 (11. ill bv tlic following svhenme.transform of periodic signals.JW) 2n 1 * (.as shown in h'ig 11. T h c spectrim of thc samplccl sigridl iii the f~rc~iiencydoniain clearly periodic with period w.. (11.33) Comvolulion of tltc spectr urn 3 (p)thc contiiiiious signal with tlie inpnlse train of in the frequencydorimin L L L ( ~ T / ~ T ) a periodic coritiiirration of 2 (p) gives at rnultiplw of 'Ln/T. I . we c'a~i it to tl ribc the sariiphg of continuorrs signals. .30) In other words: the biLsis functions of the Fourier series this into ( I 1. where we ta. The result ~ ( t )again is il. Ic E Z.7.75) t u obtain I t ( t ) = .U T x(JW) 1 dy(. Tlie freyiienry w. series of impulses. This idealisation allows us t o ideiitify the fuiidanient.3. = ' is called the sarnplzng . In Figure 11. The weight*of thc individual impulses arc' the values of t h y signal at points t = k T . Putting By cortiparing (11.jrcquency. as shown in Figure 11.ke the precise values of the voritiiiuoiw fiirictioii at saniplc points. c1 (t) and r z j t ) Ck = AkBkT.hs of the arrowb..6. (11. To rnodel idcal sampling we start with a contiiiuous signal i( C) arid nuiltiply itz t with an impulse train &LLL($).
ges of storing a digital Ggnal db oppostd t. It i s also usually desirable with most applications to rrconsl rucl the analog signal from i t s sainple t . The iimsic bigiials arr sampled het niiw there i i r c m m y tecliiiicd advantu.6: Itlcnl salripling of the signal 2 ( t ) a T AE.7: Sprctriim of the sampled function * ( t ) the) can be recorded or processed using digital tecliriology. with Compact Discs. foi exarnpl(’. Sampling anti Periodic Sigrruls  Figure 11. T I:igurc 11.270 11. a l n c s . C. but the u5rr of t h e C‘Dplayer is only .o an aiialog signal.
I nbovc n jrequency u l g .y wg 2s callcd the hand limit. wliirh we will be (wering ill t k section. wliicli irr this < bitrnrily has a triangular form for /wI < U J ~ . Sarndiiig 27 I ccl iii thc analog music sigiial.8 sliciws tlir s p d r u m of a bandlirnit~dsignal.To distiiiguish between the periodic reprtitioiis by sanipling. It is therdorc.nc. %fa bandlimited signal ? ( t ) is samplod frequcmtly enough. The frcque. so that the baseban r ~ ~ c t ~ tdoo ~ soverlap. this is dso called t h e b n w h r lpigim I I 8: Basebarid spectrum with limit frequwy "L'~ We caii USP it to formulate tbc central priiiciplr of the sampling thcorcrii. 1 F ( t ) actiially c m be r w o i i s t m c t d worth kilowing whet lirr a contiiiimus s ly what t lie sampling theorem deals from the. arid iiot llir sample values. I Figure 11.113. with. To h g i n w e i d r d i i c e the conccpt of a haiitllimited signal. F ( f ) can be interi not . samplcd hignal ~ ( f ) 'i'liis is .
f"=T=G is also called the s a ~ r ~ p l m y It specifies Iiow often the sigrral will be sampled mtc.10: Overlappiiig baseband for w g > + (11.9: For wg < $. i i m p l i n g rate is as low as p( ible. T h e ovrrlapping that occurs in the dashed regicm is also called alzaszn.9 as an obviously eirorieow reconstruction by the interpolation filter leads to highfr cquency signal coniporieiits apjxaririg at other locatiorts in the incorrectly reconstructed baseband. Sometimcs fa is called the saixqde f'requericy. . To clearly distinguish between . atid U . prr unit time. the I eriiis 'sarupling rate' and *arigtdar sarnplirlg frequency' shonlrl tw usrd. the signal X ( p ) can be reconstructed using the interpolation filter H ( p ) It this condition is violatcd. critical sampling represents the bardcr case wherc thr. 111 this the idcal intcrpolatiori filter is a iectailglc in the fr equencytlornain with frequency resporrre El(jw) arid impulse rcsporise h ( t ) (see Figure 11 11): ~ (11.10). a i d can no Ioiiger be scparatetl b y convolution (see Figure 11. Samulina and Periodic Signals xum) interpolation /filter I Figurr 11. The Figure 11.34) sampling frequency wo = + 1 W(l.1.35) . Clearly. shown in Table 11.272 11. then the repeated baseband overlaps itself. h aliasing has still becm avoided.f(. 1)iffment c4ioices of the bampling rate caiisr different cases to occur.
3.inplirrg critical sairiplirig Ny yuistfrequency undcmauiplirig =+ sliasing overlapping spectra fa = W Y 2 w!l 7r Jb i 1 Figure 11.11: Impulse resporist.12 Sampling arid iiiteipoLttior1 with a n ideal intcrpolntio~i filter In the tiirwrloorsin we can obtain with SOIIIC rcarrartgmcnt. o f ail ideal interpolation filtci Figure L 1. tlir reconstructed . Sampling 273 l&le 11.11.1: Dcwribirtg different samplc ratcs U) f > 247r oversa.
The sauipk valucs 2 ( k T ) are illerefore Ihe weights of tlw sifunctiuns. ill1 (iiifiuitely 11laily) of the siftmrtions arc superimposrd.13).shifted as sifiiiirtioris (see Figure 11. For values 1 = kT ill1 sifunctions but one are zero.iori was described by Whittaker in 1915.ible to represent i ( t ) ilcd sifimctions. Sampling and P n k d i c Signals (11. wilh (11. The in in its curr~iitform.37) Perfect reconstriictioii S ( t ) = 5(1) from the sa~nple valucs surcccds through interpolation if the saniplr values arc u ~ e d scaling factors of appropiatelj. aud iiiterpolatioii with the bifuriti.274 signal jj(t) (Figure 11. was irLtruc1iicec. . From (11. and between the sainple values. Of r o i m r .1'2) 11. Iicwever.l by Sh:tnnoii in 1.35) . This 0 1 sifiuictioii has its ~iiaxiniiim 1r value of 1 a t Illis point.. The basic idea of thc sampling thcwrm c o r i ~ s from Lagrange (17361813).(11. such that the iiearrst silrnylr vnlurs hiwe i gre:itn l influence c m the sigiial valiie t. it iiiiist be pos.948.33).han tlioilc far away.38) we can see that this requirement i s eyiiivaleiit t o handlimiting with L J ~< $.36) ( I 1.
3. wl Both f1equencies can be chosen inclepciidently. and tlic periodic c*orCinutltioii stand next t o each other at rriultiples of the sample frequency.() F[ complex. Critical sarripling is clearly always possible for complex bardpass signals. There is no condition foi the relationship betw~eeri I and Aw.14: Tiirilatrral bandpass spectrum If we sanipk this signal wiih swmplirig rate w e again obtain a periodic continuation of the spectrum whercb overlapping is orily just avoided. + X(p) + X*(jw') . they are represented by two real signals.3 San1 ling Theorem for Complex If we consider periodic continuation of thc spectrum by sampling. = A w . Figure 11. that corresponds to the sarripling frequency w.39) Ckmiplt~xbandpass signals are of special iiiiportaiice in communications. Figure I I . aiid m e for the imaginary part. The corresponding time signal must be complex as the coiiditiorrs for conjugate symmetry (9. 'Mir will soon sec that f o x an appar.U J ~ + ? ld arid +dq $. one for the real part. Figure 11. for example. This leads to the sampling theorem for complex bandpass signals. Sampliiig 275 11. we bee that tlie basebaiitl spectrum betweeit wy arid w g . Correct choice of Ihe interpolation filter for reconstriiction of Z ( t ) requires knowledge of which region ol: the frequency axis the signal occiipied before sampling. As it has 110 frequency conipontmts lower than (lwi< UJO) or higher than (iwi 2 WO +nu).it is lmown as a bundpass spectrum. To recover tile original signal.14).49) are not fulfilled: + 9 (11.rcnt1. The spectrum of lhe original signal z ( t )c o ~ ~lie just as well between . the iriterpolation filter should be chosen so that it blocks all frequcricy coiriponrrits riot in the speclrum X(JW)of the original +iial ?(t). We d a r t with a signal F ( t ) .9.Its fxeyncxncy response is a rectangle in tlie bandpass location: . with spectrum X ( J W ) (Figure 11. or even be1.ween 0 and the periodic rontiriiied spectrum X ( p ) would he f ~ 1 i ~ a ~ ~ i e n t ~ l l y indistinguishable. a (*as(& critical sampling.I5 shows the spectrum of of the sanipled sigiial and the impulse train.113.y simple case of a r e d bardpass signal that it is not.
l) Figiirc 11. Continuation belwecn wu arid o the bandpass sp I iirii witlront gaps requirvs that. w .15: Spcctrniri of the sampled complex signal (11.276 1 1. x wa = 2Aw . f u~in F'igure 11. are Crilicnl sampling rcquirei a sample rate of f " I T . so thc corresponding impiilsc response is tomplcs: (11. 1 h .16 h 1 i o ~ l . In contrast to complex banclpass signals.17 can take exactly an even ririiribcr of halfbands of width AdL1.e. correct choice of. ) for "'cl = 4Aw.40) 'I'herr is no conjugate syrriinetrp licrt. in this C R S P . I. . sthe real and iniagiiiarv parts ol / I ( . For the spparentlv simple case of a reid ban gn..il.=. eithcr.41) Figure 11. The coiiditions ol conjugate ~jmnwtry hilfillecl in this case. triticd sarnplirig is oiily possibltx if WO mid A d have i i particular rrlattionship with each other. T l i ~ of tt real badpass signal is ihowii in Figure 11 17. Sampling and Pcriodic Signals interpolation filter H(icn.sample sp~('1 iimi fiecluency is not a s simplr as for the coiiiplex bandpass s i p a l .
.16: Complex irnpulsc response of the interpolation filter Figure 11.42) Otlicrwisc thc space in Figure 11. is shifted by coil\iohitio~t that with two svmmetrical delta iinpuihes to the bandpass loca.17: Bairdpass spectrniii of a rcal signal The even iiuiriber condition comes about l m a u s r of the conjuga wo = 71 .35). Aw. Figuie 11.18 sho\’i’h the sp c callg 5arnplcd signaf ant1 the t oirwpontling dclta impulse train for n = 2. (11.tions: Thc correspoiiding irripiilse rcq)oiisct is piircly rcal iuid rescrnibles the impulse responw of thr interpolation filter (1I . but bwarise of the bendpass cliaractcr.113.17 carrriot br c*ornpletely fill m l critical sarrtplirig i s irnpossiblc.ponse in the l~as~barttl. The ideal iril crpolation filter must take exactly tlir frequency tt~iiiponent~ that were contained in the original karrdpass signal. Sampling 277 Figure 11. n cr IN . We can dehcxibe it with L rcctaiiI gular frcyienc> rec.
Aw. Sampling and Periodic Signals Figurc 11.19 agree with the curve in Figure 11.19 depicts the irnpulse respoiise for W O = 3.278 11.22).35) (Exercise 11.18: Spectrim of t.he sampled real signal it is multiplied by another cosine funct. in the lowpass case (11. It can clearly be seen that the envelopes in Figtire 11.(t) = si (g) ( cos (WO + $)t ) (1 1.1% Impulse response of art interpolation filter for a critically sampled real bandpass signal with w = 39w g The sampling frequencies deterrriiiled for the three cases of critical sampling arc .44) Figurc 11. (11.ion: h.44) is also valid for the case wg = 0.11 when the time axis is scaled by a factor of 2. Figure 11. ( n = 3 ) .
14 and 11.2: Sarripling frequencies for critical sampling basehand signal WU = 2w. We can also describe such nonideal sampling with ail inipulsc tiain. Sampling 279 shown in Table 11. cwiplex bandpass signal w.2 / 2 The integratioii time z must be smaller than the sampling interval ( z < T)..as integration: (1 1. the integrated value is only available at thv eiid of the integration interval. if we forinulate the collected charge over a time 2. In fact. the required saiq~. the sampling procedure requires that eiicsrgy is taken from the signal 5 ( t )for each sample value. 11.3. however. = Aw real bandpass signal I U. Th(1 series of rectangles that describe the sampling procedure is shown in the middle of Figure 11. The sampicd signdl ~ ( t )thus is (11. We have assumed that it is possbible to take the values i?(kT)of the sigrial . For example. They are each centralised at the desired time point.20 below). onideal Sam Until now we have Hot considered physical implementation of the sampling procedure.8.20.46) .E(t) at precisely defined points G = k T .11.17 reveals the 'L'ahle 1 I. This also requires that the spectra are cliosen so that critical sampling is possible. The resulting capacitor voltage is a measurement of the sample value. a certain tiniespan.45) f. Charging a capacitor requires. as depicted in Figure 11.20. so the capacitor voltage carinot bc assigned a definite signal value of Z ( t ) at a clearly defined point in time (see Figure 11. In this manlier we avoid having a delay term in the calculation. i s equal to the complete width of the frequeiicy baiids in which ~ ' ( J L I )is nonzero.2: Coniparisori with Figures 11.le frecpency U. =2 Ag following simple mle for choosing the sampling frequency: for critical sampling. an electrical signal can he sampled by taking some charge at each sample poiiit a i d storing it in a capacitor. although strictly speaking.
can take this iiito accwurit insing any weighting function U ( t ) with n ( t ) > 0 for . there is an additional systrrn with impulsc response h ( t ) = Lrect(L) 7 r I’ipure 11. Wt. u ( t ) is also callcd an aprrtuw f u r r c t x m .hc Tcctangle fimction.45) is iiiidtiplied by t r ( t . Coinpared to the ideal sarriplirig in Figure 11. Sampling aid Periodic Sigiials Figure 11.47) instead of t. Before the intcgration. iiieasurerrierit o a f signal value requires c2 certain time.. i ( v )in (11.20: Sampling a signal in the realworld UI‘C A diagrammatic tlrscription of tlris sampling procedure is showii irr Fig11. A fiir t t w r generalisation is the assumption that the enrrgy transfer ih not Luiiforni within thc tirnc z.< t < 2 2 a ( t ) = 0 otherwise. 7 T (11. In connection with optics.U). h i cvcry case.280 11.21 Nonidcal sarnpling of t h e signals 2(t) The dcscription of nonidcal sitinpliiig introcluced here with the c~xaiiiplc an of clcctricd sigrial also applies to other signals.21.6.
The Foririer lransforrn of (11.22 sliows a possible rprctruni of the original signal % ( j d ) (a& in Figure 11. LVr will now i i i w s t i g a t e what infltienw nortitlcal saniplirig has oil the trecliiencycli)inaiiL. Sampling ‘28L For the sampled signal.zrrimtal propcrty with icleal sampling. (1 1.48) IVe foririd llic* luriti. T T Figure 11.22: Spwtriirxi of the sampled signal In order to rstjinate the eEed of sampling O K the fxequency responsr. Figiirv 1 I .49) Comparcd with the spectrum of tlir idcal sampled signal (11. wt’ will be liniitetl to rectangular aycrtims in &c(*o~daiice (11. sampling frequency J = 2n/T. which corresporids to an integration with duration T over the errtire sampling inter\d %.46).8) and the wighting of the rectaiiglr irnpulsc A? with the rectmglr. it is striking that the spectrnm of thc original sigrial X ( p ) is wrightcd with tlie aperture M o r e perioclie corit inuation. ‘l’o siniplifiy this.33). Thc first zero of the sifimction then lies exactly at the.50) .3. the has been taken. X ( p ) ran contain wit hont ceiising aliasing ei~ors: ( 11. that thc q)( ot the sarnplcct signal consists of the corit iniiatiori of thr spcctruin of the nonsampled signal.IS) yielcls the spectruin with of the samplrd signal X(JUJ) (11.11. UT cal~ ruldtc thc value of the sifirnetion at tlie highest trc3queney that.
FOI example. impulse traiiis and approximations using short high voltage peaks arc not practical for use in electrical circuits. so because of the shift theorem. Sampling a i d Periodic Signals The highest frequency component is damped by nearly 4 dB. t h a t is shorter Lhan Cl~e sampling interval likewise reduces the damping effect. The staiicilbe functioiis can bc thought of as a delta impulse train intclpolatcd with a rectangle of width T (11. In fact. arid use instead sitmpleandhold circuits and staircase functions whose step height is the weight of lhe coir espoiiding delt. but they are riot suppressed enough Lhat t hr resultant aliasing is tolerable. econst ruction It is riot only necessary do consider sideeffects from thc irrrpleincntatioii when sanipling. rectangle irnpulse does not lie symmetrical to the sample points. components at frequencies above half the sarnple frequency are strongly daniped by interpolation.51) Figure 11. An integration time z. Real digitalanalog converters therefore do not use these signals. 1. AlthoiigIi the description of reconstruction using an impulse train (11. that blocks fregicmciw ahove half the sample frequency. In many applications this effect can be tolerated and where appropriate can be balanced out by a suitable postconnected aperture correction filler.282 11. but also during reconstruction of the continuous signals from their sample values.23 shows an impulse train and the corresponding reclangle function. If hiirh frequency components c r cur they still should be suppressed by a special antialiasing prefilter . for all lower fiequencies the damping is lower. s .36 t o 11.38) is theoretically very elegant. instead I Figure 1 1 23: Signal interpolated with a sampleandhold circuit shifted from the symmetrical position by T / Z . most ‘1’V cameras hm7e a n aperture correctiori filter that compensates for the filtering effect of the row sampling of the image. The. We still keep the advarit ages of the delta impulse train.n inipulw and SO corresponds t o the actual sample vahie. however.
This damps tlie reorcurienres of the bascbaiid (shaded in Figure 11. for example. "T . T __ T I'igure 11.2 The problems described with reconstruction also occur in film projections. 283 (11.24).JJ). he audible highfrequency noise in an audio signal. response h(t) (1i 35).11. images are t a b i i .. so the sampling frequency is f I .52) Figure 11.24: Magnitude response l X ' ( j w ) l of tlir signal irtterpolated with sampleadhold circuit Example 11.d o r ~ ~ i i Per second. Saniplirig corresponding exponential term appears in the spectrum.While in Figure 11. 24 individid in.22 the hascband sycctrum was weighted with a silunction and then periodicidly continued. 'She shaded area could. The residual reduction in highfrequency components in tlie baseband can likewise by coinpensated for using a filter which can 1~ iinplemc'rited either digitally before the DAC or after as an analog filter. tliese noise components can be fully eliminat8ecl postconnecting a imvpass filter with by iunpulsc.24 shows the magnitude of the resulting spectrum X'(~.3. here thc periodic spectrum X ( j w ) is already provided arid is then weighted by niultiplyirig it with the sifunction. but unfortunately does not fully suppress thern. However.= 24B~. or appear in an imtzge as visibly 'blocky' pixel5. Recording and replaying it srries of individual images can be considered as sampling of a scme continuously changiizg in the t i ~ e . unlike ahsiiig errors which cannot be separated from the desired signal.
It doubl~s h t arriple rate (to 48 HE). IVe reprrserit the doublr projcction of a11 image as the ilupulsc>responsc of x i inte~~)olati(~ri (see Figure 11. Tlic fiicker up t o around 60 Hz. There is no pr:wtical interpolatiori filter available foi this optical signal. llicly ale riot there for thc eritire interval T = 1/248. inkvupting i s done with a niechanical hlirrtl which i s easily 1 0 implcmcnt.~inctlt2icorc. but tlie first reoccuiieiice at 24Hz .284 11.tically.9.55) . so the reoccurrences of the spectrum at 24 are iiot removcd and w e p m c e i v d by the eyes as Aiclcr. and pairs of coriseqitential sample valiics are equal because they c ~ r r i e from the same original. altliougli in fa(%. To avoid this problem.53) (11. a trick can be used: iristeittl of interrupting on of an image once for tlie filiri transport. as hc%wen two images. SarnDling arid Periodic Signals Interpolation njlicii the film is played back is v ~ a 5 y iiiiplrmtmt w hell each to image is projected for the corresponding lcngth iine. it is interrupted twice (see 5). the film has to l w moved. ‘ L movement of film The effectiveiicss of the clouble projcction proccdurc can be cxp1. as in Figuie 11. The.25) and find the spectrum: filter (11.
We can rxpress this matherna. Signals repeated in the timedomain may not overlap.26 shows the periodic spectruni correspondilzg t o 24 irriagPs per second and the weighting by the interpolation functiori h ( t ) . An image aigiial has a doniiizant frequency component at w = 0. Tlie zero of the spectrnm at the sariipling frrquency 24 Hz extensively suppresses the first reoccurrence.3.11.58) . Sainpling 285 Figure 11. the flicker reduction only functions when both rec that form h ( t ) (Figure 11. zero at 24 Hz suppresses annoying flicker 24Hz 48Hz U) 2.24).26: Magoitudc response of the interpolation f i h r with a zero at 24 Ilz equencyDomain The sampling theorem states that bandlimited signals can be reconstructed froin their sample values if the samptixrg frequency is high enough thal 110 overlappiiig occurs in the frec~i~cncydomain. a dual of thc sariipliiig theorem can be foirriulated: Signals of finite duration can be unambiguously reconstructed from a sufficient number of samples from their spectrum. Additionally.z Figure 11.25) a w shiftccl away from each other by exactly T/2.tically with the transform pair Y(jw) = I I Y(JW) U 2n (c)  (I157) + (11. hit the flickering becomes much 1 ticeable. weakened significantly. Becaiisc of the duality between the timedomain and frcqucncydomain. Reduction of the compoiients at both 24 Hz arid 48 Hz (*an ben be achicvetl using 1 ion (Excrcisc 11. The second at 48 Hz is not.
xercises xereise 11. 'rhen y ( t ) ca. Exercise 11. achieved by splitting the signel into sections with finite duration. an errorfree reconstruction is impossible. In this context we use the term tzm~domnza &using. Saarrpling in the frequencydomain is always carried out if values of a. In the frequencydomain.286 11.h a delta impulse train. that transforms the line spectrum into the smooth spectrum belonging to S i t ) .3 Coiisider a delta impulse train shifted by to: . that causes periodic continuation of $(t) in the timedomain.58). The relationship bctween a series of finite length and their spectra is represented by the discrete E'ourier tmrisfo~m.he frequencydomain is again idealized in (11. sampling in the frequencydomain does riot lead to overlapping in the timedomain. (DFT).12) i. for example.his correspoiids bo a11 interpolation (convolution) with the spectrum si (uT/2). spectrum are t o be stored in t. 5 0 5 10 t Exercise 11. This corresponds to convoliitioii (11. The necessary restvict. Note: use tmnsform pair (11. where continuous functions cannot exist. oiily a series of numbers. Sampling and Pcriodic Signals Sanipling iii t.2 Determine the Foiirier transform X ( j w ) of ~ ( k )= UL(at) and sketch ~ ( tand ) X (jw)for a = a == 1 and U = 3.57) as mult. If the signal G(t) is limited in time and its duration is less than T .he memory of a.n be recovered from y ( t ) by multiplication with a time window of leng%liT . If overlapping occurs in the timedomain. cornpiiter. t.i~~licati(j~i wit.1 Write the sketched signal with a sum of delta impulses as well a with the _UL ? symbol.iorito timelimited functions is.
.Write X(gw) as a S I of ~delta impulses. b) What syniinetiy doe5 x ( t ) have for t(i = 0 . For the sketch.(/) = cos5((2qt) x siri(wot) c) 2. a) x C J ( l= cos(3wot) x siri'((2wot) ) b) . b) Give the Fouricr transforms Xl(jw) . c) Sketch X ( p ) for to = and = 4.0 X I (p) zZ(t)o and the Usymbol and skel tli both functions of time. arid give the periods where possible. Note: represe~itX..( t )= sin*(3wot) x cs(0) o'wt a. and X ~ ( j w ) . . Note: see symrrwtries (9.) Test which arc periodic. t o = f and to = $'? Use your illismrs to derive the symmetry af X ( J W ) .*.Q. reprtwrit the result as a sum of deltairiipulses. Note: evaluating the coefficient forrriula is not necessary.jw) and X Z ( j w )as well as r l ( f )0.(jw) as a sum of Iwo spectra. Exercises a) Give c(t) and X(jw) witli 287 _uI hymbols. Exercise 11. Exercise 11.5 Calculate the Fourier scrics of t hr following prriodic functioiis using a siiitable analysis. XI 0'64 X2CjQ)) .11 A.4 Give Xl(.61). E t . . .* 1 x w 3 0 1 2 3 6 9 (0 E .
10 Give thcFouricr transiiorm of ~ ( t )rect = z ( t )x LLI Calculate X. 1 r ~periodic continuation n.. a) Calctila.. see (11. HOWdo the weights of Ihe delta inipulws of X ( p ) tic in with X ( p ) ? b) Givc the Fourier series expansion of s ( f ) with (11. Exercise 11. the fundnruental angular iireyuency W O for Give each.288 11.lrat X ( p ) = IL f~ J C c w =U ..18) and skctch the series of the Fourier rocffiricrits A.(JW) (t).te and sketch X ( p ) and their X(jrw).8 Shotv t. Exercise 11. in a corriplex Fouricr series x ( i ) = A. Sampling and Periodic Sigiials Exercise 11.7). P Exercise 11.p(t) = a) using coiivolution in the timedoamin.7 Expand the following periodic signals using the cueffirierit formula.~JJ"O/'~. (5) and t . by expanding tlie riglit hanct term into ii Fouricr series.9 Consider 2(t> si2(nt) arid the periodic signal s(t>= ~ ( * ) t :U (2). b) iising the wiivoliitiozi theoreai. .
and are two posbihle 10 risforiri on ( 11. x(f) t T T c Exercise 11.17).7. rcctangle signal x j t ) .11. First give the t*oiinection between wu arid T. Derive the relationship for general 1)erintls T .1 to bc expr A. t h a t thr c y ~ l i c so c*onr.oefficie to the soliition: a) starting with (11.14) arid (11.14 C~lciililtr1 1 1 ~cyclic convolution of the signals {(tj and g ( t ) = siri(~gt)using I4xirit.. .15 Consider the spectrum X ( p ) of the sigiial n:(t).r series.18). . Find the output signal : / ( t )as a Foiirier series. usiiig the results froiii Exercise 11.11 (11. f(t> * * ( I  t Exercise 11. I Find the Fourier tianqform of (he signal froin Exercihe 1 1.4.13 A system with impiilsc rcAsponsc h ( t ) = siii(l) CJ" I t c ( t ) is excited by a pclriodit. There its Fouriclr c.22) allows [lie spectiiirn of a sigrial J ( t )with period CI' .ohitioxiis defined properly.2.3. h) using tlrc Fourie Exercise 11 12 . Exercises 289 Exercise 11. Sec Section 11.
which cases does aliasing occur? For 271.290 11.) Whaf is the spectrum R ( j w ) = F ( r ( t ) }of the sigial for hoth ixra.')= WO u=m R(JW1) 6(w .2 What rases arc critically Exercise 11.Jm). 3w!? The sketched rectangular signal r(t) will be inriestigat ed 4n 1a i a. c) Slcetcli the hictioiis of time r . v E z l o form a signal Z A ( ~ )of weighted delta impulses. Sampling and Periodic Signals and Xi = sa mplcd'! 8. b). a) Give the origirial spectrurn X ( j 0 j ) = F{z(L))(sketxh with Iabelled axes). corrcspondirig to the eases t v ) . . f i ) .This cieates the ilew sprctmm: x Ru(3. where the weighted delta iiiipulses arc rcpreseiitrtl by arrows of rorrwpondilig length.1. Y) WO = F.7w) is saiiiplcd at equidistant points spaced by uo. y ) in part. b) Sketch tlie spcx:trimi X n ( j w ) a Exercise 11.gnitude and phase? Sketch i R ( j w )1. . Sketch the magnitude of this new spectrum for ll 5 w $ with a) w* = .trum R(.16 A sigiiad ~ ( t )hi = wq (y ) is sampled at eqniciistant points in time i/T.x 47'' 27r 4) WO = fl. 1)) The spec. ( f ) .17 o Z A ( ~ )for 2n the case T = .
Sketch X.c. c) Sketch Xa(7w‘)for sampling with f a = Exercise 11.  e~~l’clt..I I .?w) the output signal.r..105 a) Sketch the spectrum Y(. .. complex modulation XO’O.19 2. of b) Draw a 1 arrmngement of sampler and reconslrnctioir filter N ( p ) that caii 1 deniodiil&e y ( t ) . What are the possible sampling lrequeucies? S1rt. . = coo . and wY as above.) = Sn and for wy = 2n. so that x ( t ) is obtained at the output. :symmetry docs ~ ( t ) have‘! b) The signal x c l ( t ) created by critical sampling of ~ ( t ) . WO The signal s(f)with the sketched spectiinn X ( j w ) is rriotliilated with the shown a r r angrinent . Exercises 291 a) 1s x ( f ) rt.tc.20 A signal with the sketched spectrum is sampled ideally w k h intervals T 1XO’to) .) 1 . the sampling is Give frequency and the sample interval 7.?W) for U. for this c .al\iaiucd? What.. 4 . Exercise 11.Jw) tor a suitable reconstruction I fil1.h I H(.(.
. f r c y ~ ~ f(.a) Is t h e sigiral a highpass.23 A signal ~ ( t )bandlimited to f . ~~ The nonideal proprrties of t h sampling are represerrted by the apertme f L i n C t i o K i ~ t 13. or a btliidpass slgrlitl? Is it ronqlex or h) Is rri1.O 7 a(t)= 1 t z 0 for O < ( t < T olherwisc . c y .2 x.a~~ (1 1.tl sarnpling possible'! c ) I)raw the spectnmi X . .ic. = $.I lowpass. real? . case Exercise 11.35) for " ~ ' ( 1T 0.T z r. ( j w ) o€the sarriplcd signal lor T .22 Show that the irnpulsc.44) 1)ecornesthe 1owp. Wli 8ancl ie') Exercise 11. response of the ideal iriterpolalioii filler for ciitically sampled real bantipass sigrials (11. = 0.. = '20 IrHz is vritically sampled at.for r<t<.
as clescribed in (11 5 3 ) . Then H ( p ) and skctch J H ( j u ) l€01 T = In particular.i)9 0 a ( t ) and a suitable example for X ( p ) in tlsc same slietch. &. Exercise 11.a f = 10 kHL.24 The effect of triple projection 011 the flirlterivlg of cinematic films (see Examplc 11. try to deiitify the zeros of H ( p ) . The interpolation filtw takes the form of a mechanical sliu with t he sketched Express h ( t ) using reclhurtions and (5impnlses.ed. Exerrises 293 a) Draw A(j. less at the I m d )~ limits? c ) DctPrmine z so that the aperture functioii riialirs il suitahle prcliniiriary anti aliasing filter if sarripling is carried out at .11. t ~ rnaiiy dB is it. b) Comparc the amplification of the aperture fL1nction at the frrqurircy zero and at the band limits of 1ht3 signal. . By 1 .4.2) will br imestigat.
This Page Intentionally Left Blank .
$fmm.3 and 1.14. W e need iiew tools to descrihe such secpien . This traiisforni is also rcferrcd . which we will iisc to exaniirie discrete signals in the frcqueiicydomain.1 shows the conventions we will use to reprrsent such signals.In Chapters 1to 10 we gcjt to know wmc powertul tools for dealirig with contiiiuous signals arid systems We carried out the convei sion of redworld coritiiiiioirs signals t o sanipled signals in Chapter 1 I. or the pixel address within P~C a digital image. could A l l br iised.11 A discretetime signal is represented by a sequence of iiuriihcrs that is callrd a time series. Examples of such an index might be the S ~ ~ I T ~nuniber fox a digital aiidio signal. we put the independent vaf iable in square brackt. There is no sniooth transition between the nurnljers. the zlrunsform. In a coiiipiater. The discretetime Fowler transform will also be introduced. This chapter dcals with dzscrrte signulr and the discrete form of the Foiirier transtorm. to as dhwretetme Fourqrr traiLcsfOrm (DTFT). of these are represented in Figurc~1. Figure 12. that are defined over a discrete range by a numrrical value (the index). the J trun. a cliscretetime signal arises from the sarnplzrry of a continuoiistiiii~ signal 5 ( i ) . We will sec that i t has sirriilai properties to the Foiarier transform for coiitinuous sign&. we can only work with a sequence of niuriberi. At the end of this chapter we will investigate the relationship between coiitiiiiioiis signals and their disrrete cquivaleiit as a series of samples. In addition there are discrete sigrials t h t haw not arisen from sampling a continuous signal. The sampled signals were tieated as conitiniioiistiinevariables so thal the rools we had learnt . The two following (liaptrrs deal with dzscretr systems and we will leain the discrete counterpart to the Lsplace transform . which is necessary foi digital processing. wherc a sample is taken kern P ( t ) at regular . arid w e will concentrate or1 them in Chsptcrs 12 . Example:. 42. like the Foiirier transforin.4. in order to distinguish them from continuoustime signal$. In Swtions 12.1 arid 12 2 wtl will coiisider diicrete signals together with sonic exaniples. In inany technical appiications.
X.s a numerical value :r[ k ] : . this 1s not the case if the series is going to bc proce a computer. they could also be conipltx r [ k ] E C. called quantrsutzon. although ~ ~ (is1zero tor almost all of these timepoints (except for C = AT. so intcgration of is ~ [ k i ] not possible. The Spectrum of Discrete Signals inrervads T ant1 is translated 8. The finite word length means that numbers cari only be within thc limits of tlie nurriber representation being i i s ~ d . A cornputer with sufficiently great word length can work with digital signals that are so finely quanl i d . Thc dependent variable f is tfefincd for every point in IR. ~ [ k ] signals tlien become themselves discrete. how(\ver. E Z). we refei to thcsc signals as d291td szgnals. 'Vo distinguish tlieiri fIoin the di with coiit inuoiis values.1: R. ) In contrast. in arl w e will consider general tirnriseries r[k:. kG z. Striclly spcaking. The values x [ k ] Figure 12. Mie will therefore restrict ourselves to discrete signals from 1 1 0 ~ 011.. s[k] only defined where the index k is rtri integer. . We will return to rising sampling in Section 12 4. Rounding to discrc vtzlues i s a nonlinear proc Processes involving digital signals therefore cannot>be described by LTIsystems.1) Here it i s essential t o distinguish the discrete signal z[kj from the sampled signal introduced in Chapter 11. and the Fomier transform trom Chapter 9 cannot be usecl.and the values .atioriof a discretetirnc signal :c[k] are t*hemselvescontinuous and in geneial.epresent. that they call approximate discrete signals with contiriiious values. a n d i.T).296 12. s T the following sectioiis wc will deal with discrete signals in depth without being n lirriitecl to sqnences of sainplecl values. (12.E[q = %(X.
2 shows thib series.lw contiinioiis deka impulse: the unit. irnpulse 6jk] is not a distribution. In contrast to thcl [] continuous tlelta irripiilse 6 ( t ) . u r r i t impulse Comparing (.4).ion of a continuous sigrlal with summat ion often leads to the equivalent discrete signal. simple. h continuoiihtime delta impulse ~ shows a close relatioiiship between the two: (12. Thcre is a m o l e \ significant. in (12. diffcrence bet8weeri disrrete nnit impulse arid ~ I I P t.5) correspoiids 1 o smiiniatioii with respect to tJie di ete inrlcpcirdeiit variwhle K . r i z ~ ) easily checked 1 ~ k Figure 12. The discrele unit impulse Iiah the helertiw propel t y " U I (12.5) Integration with resp to tlie continuous variable z in (12.3) Figure 12. TVe will sec that replacing integration in the clefinit. expoiieiitial series. which makes calculations. x w Itnow thrir continuoustinit.4) that. ilie unit impulse a forriitila.2:Discrc4t. 'L'heso sigrials are the inipnlsc. ~ 1 The discrete unit imptllse is defined by (L2.hiswith the relwtivc property of f . coiirti carparts. and does not wquirc the use of the SP .aling vith some simple disrrtlte siguals. For k = 0 it has the finite valiie 6 0 = I .
.he angular ! frequency. k Figure 12.2.6) The relationship betwcwi the nnit step function arid the unit impidlse corresponds to the tquivalcrit for continuous signals: (12. The Spectrum of Discrete Signals iscrete Step finction The discrete unit step function from Figure 12.Elk  1). iinpiihe K ~ S ~ W Iof a W discrct e system.9) X is the complex amplitude.3 is defined by (12. arid K is t.298 12. Discrete cxporieiitial sequences axe often characterised directly by giviiig theii base. especially if they will be initerprcted a s tlx. I (12.3: Discrete unit step function 2.8) Iiit.3 Exponential Sequences ExponenLial aegiwnces are in general complex sequences of the form (12.7) I 6 [ k ]= E[k] .egration in the coiitiriuoiis cabe corresponds bo discrete surnmation and differentiation of coritinuous signals corresponds to subtracting the neighbouring valr L e in the discrete case. C is the damping (w1stitnt.
4.12) .4: Examples of expctrieril.11).1112. arid illustrated in Figure 12.11) For b o t ~ exponential series X = I and 1 = .2. adding multiples of 2n to 0 does not change the sequence: p k e j(62L27r)k (12.12. In contrast to continuous mqmiicntial fuiictioiis. tiiffcwnt values of tlre nngitlsr frequency R can lead to the hame exponential series.1 le T w o exainples for uiiilat. As eJRk is only etaluated for integers k . Some Siinple Seqiiences 299 1 1 2 Figure 12.10) (12. ( 12.eral exponential sequiices with real values are given by (12.ial secperices E x ~ ~ p 12. Both srries are the same except the 2 signs before the odd valucts. TIX arigular frequency II ~ i a s i : the values 5 = 0 and I2 = n respectively.10) and (12.
o 1x3 turning biLclmmrds.2 The expoiicutial function in Figure 12.5: Exartiples of complex t~porieuitialstxitas Example 12. ofthe danrpirig constant c = 0. it is iiecessary to limit the r~recluericvtlomairi the to width 2 ~ .12): tlie uppcr curve has been calwlated with ari angular fr. 33ecaiisr of this rffect. Incrcwing k by 1 turns the v d i i r ~of the scries by 0 . i i o ~ e v ci~ecausc ~.eqiiency of ( I = 0 . where samplirrg causes different freqiiencies t v oveilap. 1 racl less 1 hair a whole circle.27i.300 12.97i.he iriiagiuary part). The exgonenliwl series with aiigiilar freqiicnry 12 = 1.9. R lower frequency.32). tlir absoliitc value is iednced to exp(O. the Iowrr c i i ~ v e does not seem to have a higher freqiioricy.l x 10) = 1/e. 2 ~a d in the. iti identical to the exponeiittinl quency Q = 1. . TEic Spectriirn of Discrete Sigrittls 4 Re ( x l k ]) Figure 12. f i is only 0 . The series rnakes almost 1 one full t u n each time. 2 ~ . . For l k = 10 Ihr samc direction as for h = 0 is Icyeated lwcniise 10sl .171. The ambiguousness of the fxquency o f discrete exponential hinctions is t h c reason for the oct'urrcnt'e of 'Aliasing' (compare Chapter 11.2~ = 0. but iiistead.5 illustrates thr anibiguouu charactflu of (12. r mathematicnlly positive direction (from the r ~ a part t o f. 1 1the lower cwve the angalar freqiientyis $ 1 = 1. for it spectral represcntatiori of di'sci ete signals. e '\ already know this plieiioinerioii n movies. wlrcn the wheels 01 a wagon rtppcar t.1. In comparison ~ to the upper CUTVF".9. t ttiriirig in the oppositc dircctiori.
the discretetime Fourier lransfvrrri (DTFT). DiscreteTime Fourier Transform.1) introduced in Chapt. its ust in the frequencydornain requires an inverse transform. It consisls of air inlegration of ~ ( r 7 ~ ~ ) t m r one period of the spectriun (12. il is periodic wit li 2n. the sum of the series s [ k ] is finite: (12.3.he discretetzme Fourzer transform or the transform as: (12. t r aiisform pairs. iscreteTime Fourier As a series z [ k ]is only dcfincd for discrete values of k E Z. * 1 E transform arid d&nc the Fourier transform over the unity circle of the complex plane. X ( e J " ) is also called the speetrum of U senes.we cannot use the Fourier iritegral (9.. As with the Fourier and Laplace transforms. We therefore ctefine t.r and tlic Fourier coeficieiits arp the valu f In order to rccover the series s [ k ]Goni the spectrum X ( e ? " ) .13) rcpreserrts a Fuurier series o X(e'"'). A sufw~errt corrditzn?~to shorn7 the existence of the spc rum . we haw t o iise the formula for finding Foiirier coefficients.14) . theorems and symmetry properties. This coiivcntion will make Ihe transfer t o the ztransforiii easier. so X ( e 3 ( $ 2 + 2 q= SjeJ").r[X]into a coritimious complex fiinct?ionof a real variable $2. as each term of the sum in (12.er I).he .13) contains a 27iperiodie term i27k'L 11 order to see this more cleaily. In contrast to a coiilinuoiis signal.7== { x [ k ] is that ). (12. The period is 2.13) e ~ ~ i tof the i o ~ It transforms a series .16) I t ~ ~ g 1 This relat iorisliip represents the znverse dzso ehetime Pbnnw transforni . w~ miritc elf' as the argument of l.12.15) isereteTime The defiriitioii of the spectrum of a sequence from (12. 30 1 e It wonld be comwiierit to iise the advantages oflooking a t continuoustime signals in the frequencyclomain with ciiscrrtethe signa To this end we ad1 be introducing t h r discrete counterpart 1o the Fourier transform.
3. (12. * Transform of an Undamped Complex Exponential Series F e3"Ok.1 +F*Transform of a iserete Unit Impulse 'li.J $ Z h . Thcse were the uriit impulse.21) fc U (.17) We thus obtain the transform pair (12 18) I T h e transform of a shifted unit impiilse .19) and so the transform pair is We can PP that.3.L[/C) of phase as for a contiiiuoustime signal: Ark . To find the F*transform of an uiidaniped bilateral exponcntial scries start wit. The Sgec*trumof Discrete Signals iscreteTime Fonrier This section containb the Fourier transforms of the simple sequences discussed in Section 12. As a concluding example we will deterrilirre the spectrum of a rectangle ftmction. 3.hat we had derived from equation (11. and insert the unit impulse.18) is B special rase for K = 0.C. Now we are using it to determine the spcctruni of . transform.K ] leads to a linear change = e. The result caii he given immediately. . tbe transform pAr (12.fFje3SZk k (12. calculate the 7% traiisform of a unit iiiipirlse x [ k ] = 6 [ k ]we start with the defining equation of the . q/C.302 12.r[k]= eJnrlk: .2 . unit stcp tunction and bilateral exponential series.2.h the interesting relationship we (12. using the selcetive property of the unit impiilse: X(&) = CG[kje'~~k I = k .7) in thc last clmpter.
25) (12.E*[k . (12.23) can also 17e confirmed simply by using the inverse F. v E 22.23). t ransforrn on LLL A special case ot (12.27) [ k ] can be obtainetl directly The Fortrier transfclrm of the constanl term from (12.&a[k] . The ambiguousness we discwsed in Section 12. 12.3 likewise applies to I the spect rum of tlie delta impulse at (1 = 2nw. The rwulting impulse train in the frequencyspectrurri c m be elegantly represented using the shasymbol we idroducetl in Chapter 11. 3o < k < x 2 a n d a bilateral step series with no middle value it constant (12.3 F.3. DiscreteTime Fourier Trarislorrri The trnnsform pair is 303 and states that tlie spectrum of an rxponential series eJCiok in fa( is at L = Ro. The correctness of the transform pair (12. (1 2. the transformation yields + ( ) 9 .26) E[k]= E l !k]I.12.29) SubBtitxting (12.26) confirms this.23) is the spectmm of the sequence ~ [ k= 1. I ] f = 0 is put into (12.2.24) EI ( 12.3.Transform of the Discrete Unit Step F'unction Thc discrete mit step function ~ [ kcan be expressed a the sum of ] s term 1 E l [ k j = .11.28) ] To determine the Fourier transforiii of the second term ~ 2 [ kwe express the uiiit iriipulse drk] by &2[k] spj I = Q [ k ]..3. .
be eviclewit why we split ~ [ kup into cl[k] i ] ~zlA1.E* Transform of a Unilateral ~ x ~ o n ~ n tSequence ial \VP will now find the . { ~ 2 r ~ ] \ (= 1 ~ and from here we see that (12. 0. thc traiisfcr function of a n accumulator.32) togetlier. kiowrmi. { E ~ [.heiJ obtain frorn (12.32) can only he used for 52 # . .e]. With thc . We also foiind two tPriris there: one delta impulse and a term for wliirh s = J W citrne from the transfer function of ari integiator. we finally obtain from (12.r If . ( 123 0 ) together with (12.5.l ] )  I]} = F*{EZ[k]) e3"3*{52[k.92).4 .31) As wc cannot dividr by zero.18) we t.F*{Ea[k]} F . aiid ~ 7 "will show that it iii fact rtpreseiils the discrete counterpart of a i l integrator (sce Example 14. .33) We now con?pare this result with the Iiourier trarisform of the unit step functioii in (9.26).3. 12. By adding (12.?=*{E~ [ k ] } would contain delta impulses at these fIecpeucies thcy would havt. we obtain F*{&'$T] & 2 [ k . to be considered separately.. The Spectrum of Discrete Signals Using the laws of Iinparity and displacement from Section 12.304 12.7}6 2 ~ 3 . (12. 2n.j).. transform of a iiiiilateral expoiieiltial sequence x[k! = I ? E [ ~ ] with a EC from the defining eqrration (12. Tlie other trrrri happens to I)t.28) and (12. It should noxi. U c Z.L. 5 2 [ k ]has zero wean and therefore there can be no delta impulses at 62 = 27w. * transform.F*{E2[k =.27r. F the discrete unit step fiiriction produces an inipulsc train instead of a11indivichial impulse. 4.34) We know that thc sum of an infinite gcwrietrir series is c2 7 (12.29) .3.35) .27) (12. With (12.13): (12.5).
liriitc geometr ic wriw (12.rwnsform ~ o F x ( P of a rrctanglc scries wit11 length N ) ( 123 7 ) as showir in Figure 12.12.6 shoivs 011 the lelt the first few vdues of this fqxmcritiwl series for it real (1. T h e spedrurri is clearly periodit with period 2n.38) .36) does iiot convc'rgt' for (1 = I . the sprc'trim of the [uri fimrtioii (12. DiscreteTimc Foiirirr Transform m t l from t h i s TVP 305 can dircctly obtain Figure 12.36) ab a spc 1 ansform Qf a Rectangle Series ~ + w:mt to find the . As (12.3.7. and the right shows thc magnitude of the spectrum.33) is iiot contaiiied in (12. Inserting this inlo tlir tic4ining cquatiorl (12 13) Icads to N1 Figixrct 12 7: Rectangle serirs of length N a. * t.
(12.8 Figure 12.c[k]of the discrete time signal (see (12. According to this simple relationship in the timcdomain.306 that can be rearranged as 12.1)) (12. The sigiial .41) . that have tome about.9) will first of all be represented by ifii impulse train Ir(kT)S(t. as in (12. from the sampling of continuoustime signals 2(L). that mcaris bvtwcen the Fouricr transform of t h r sampled continuoustime signal and the spectrum of the discretetime signal. wc have discussed seyitences in gerternl. a corresponding relationship in the frec~iimcydornaiii shotiltl exist.8: Magnitiide spectrum of a rectangle scries for clifevent IV Up t o IIOW.k T ) .oiitimioustimr + p a l F ( t ) with ideal s. Kow we will examine wqucnccs x i k ] . ( t )tlerived from the c. The ~ intiividuaI members of a series xjkl are h e sarnpld valurs of t h continuoustime signal B ( k T ) . The Spectrum of Uiscretc Signals The resull is shown in Figure 12.timpliiig (see Figure 12.40) W e expiess t h e weighting of the delta impulses by the values .1). without consicleririg where they come lrorn or what they represent.T.
io~~ship between the 7 and 3. ( 1 ) is the saim as the spectrum of' (tic discrete signal . thr sim for ern.ausc of the sampling pi0 . if It = u!T is understood to be c. rrorrrr&rtl wit 11 .12.&i. lhe series of: sampled vahies ~ [ k ] the i~iid tlicir spectra arc tlepicled iri Figure 12. spectrum The relat ioiiship lictcvecii F arid Fk spectra gives a n irnpor l m t insight that allows us to transfe'er rnmy imp per ties aiitl thwrenis th.jarxiplc ixitexval T .1 ) yields 30'7 (12.ontiiinoustizrrc signal x.13): that the spectra agree.4..13) The periodic Fom ie1 traiisforrrr of the.Tlie cliinensionlcss angular frequency it of ~ ( r . sariqilcd c. spectra of rontinuoiistime sigria a of discretc(iirie signals. continuous signals discrete signals . miinot he applicd to discret c>tiriie sigiids bet.9. In the following section tlie niobt important givrii It is casy to recognis.r[k].42) k We see fk)iil cornparisorr with the tPe:firlitiort of thc: F*transforln ( I 2.' "WII ) of t h r angular frequency of X ( cl"). c a ( f ) .1): R.\.e t h a t soixrr th~urcnis. emiiiplv.he normalised arigtdar freqiiency: ( I 2.oiitiriiioiistimc sigiia\ . Sayripling Continuous Signals Transfoming xLl(t) with ttra E'ouricr integral (9. The rclat iondiips h t w w n the cwntinuoustinrc signal ?i'(t).it) Figure l2.it apply t o the. sampled 1.
.IZrorii thr linmrit) of tlir siiinn ion in (12. t r F of integration.13). it. follows dirwtlv that tile prirrclple osition still applics to the . tlic sciinc i s true for the i (12.46) .
ins R y insert irig it into ( I2 IS).6. ii c a n b~ shown (see Excrcisr 12.9) t h n l ~ ( 12. in contieclion with diacrcto WIsyhi <.48) .This will hc nrore accurately disc iissed in Chap(er 14.
erties of the sier \Ve definc even aiicl odd sequences so that.$lit:nd~ior~ (12.51) (12.52) Accordingly.S A.f[k]. T h e Soectrurn of Discrete Signals The Parseval rel:la. c m also be formd by iiii (>grating I' with respect to lF(e'")12 in tlw frecluencpdornain.55) .54) Adding both equations confirms that: Z g [ k ]4r u [ k ] z [ k ] . the symmetry axis goc~. the Parseval relatioiiship st iltes that the energy of ii time sipital. for odd sequences zu[Ol= 0.50) As with contiimonstimr sigiials.310 12. dcfiiied hrrr by \umniation of I. Every sequence ~ [ kcan be split iiito ] an even tind an ocld part: (12. special casc of t.53) (12.tionshipc a n be obtained property with g [ k ]= f"[k:]: H. exactly tliroiigli the clement 2[01: Even sequences Ckld set~ucllccs sY[k]= a'Y[k] xu [k]= II' I . [ k ] (12. = (12.he mul~.
.59) 7 he spcctrum o a series with real valiirs is therefore real for f i = .20). (12. if R = T . for example. . . (12.39) 2 ~ confirm this. % F Transform 3J 1 Exarnpla 12. .33) niid (12... 3 7 .4 We have iilreatiy rrrct the trarrsforni pair (12.12. .. Propcrties of the .korn (12.~ .58) cim only fuliill tliese coriclitions. In p a r t i d a r .3 ~.5. wc' expect conjugate synimetrical spectra for discrtitctime signals with real values. 7 r . .24): I. The Lransform pairs (12.56).. for i ~ n ( X ( e J ' ~) O ) Li I (12 .
161 b y iriseriing thc defining cquation of the F .312 12.13). transforrrt (12.C. * transform (12. . Tlrc Yrx~trurnof Discrctc Siena15 xercise 12.3 Vmify the invc3rsc .
.
. 2 3 ) by rising the inwrsc F*transforin 0x1 iii (v).48).. Exercise 12.9 Show t h convolution theorem (12.8 S~LCITV by using the deliriirrg eqiiation (12. Y I ~ Exercise 12. Exercise 12. 1)) the xnuciulation thetwxn.13) of the F. .13) of the F*l.16).10 Show the multiplication tli~orem 12. liking the cyclic coniiwlrrtion irit. a) thc shift theorem.nsform.rc?.egritl am1 ( inverse traiisformation ot the result with (12. You will need t o use the convolution ~ for f[k]* ylk] arid the clcfiiiiiig cquatioii (12.7 Show (1 2 .Exer~ise 12.49j.
transform a n d tlw Foiirirr. ancl the properlies of the ztransform and i m m w ztrmn4orm.r[k]whew xi < k < oc.15)).)t>tween the . It.variahlc s to t..r(t). For continuoustime signals lioxvcvc1. I ) in two ways: by coniparison with the Laure i i ftiiiction of A coinplex nrgiiiricnl. whirh inay have comp1t. transf o ~ m and then the relationship betwc. ier transform J.{.cn the ztransform arid the Laplace txans. we first. form.c[k]) froin ( xrri for discrt. the rcgioii of coaxer gelice.x s.wkiich assigns a fiinction X(.In Chapter 12 w e got to know cii. It is (clearly) iiot named a f t e ~R fariions I riormally iisrd for i t s tortiplex frequmcy inathcmatici.n we tiiscussrti the Laplace transfotorrri.cc trmsform L{. we recognise t h a l the> \dues of the seyimiw 1 :k] represent t hc cocfficicnts of tlie . in v.1) usuaIly oiiIy twnvcrges for certain vahtes of 2 . 2 Its discussion in this chapter will deal with t l r ~ same topics as in Cliap~er 4.r.r'.transform's Latnent series at the .tctinie signals and the Fourier t r aiisfornr F{ ( f ) } continuousi ime signals taii be exp~ewvi. w1it.x ~ l ~ n i c n t b y a comp1t.or(t) of thc complcx fiequenq.r(t)}.hr tirncsigrial . but.~ri. W t b c m thiiik of (1 3. of all firid tlie rrlationship I.43). (see (4.tblr: .wrsformation for discrete(imr) signals is the 2transform. From t lie defiriition of the =transform. A c~omparablrt.. After that. The irrfiriite sum in (13. r for the rc~lal ionship (12. i t i n c t ion X(2) iri tlie cornplcx rplane. w r consider coiiwrgcnc.s). The general ciefiriition of the 2transform call be used with a bilateral sequence . we also know the Ltq&. in I t reprcwnts a scqiierice n [ k ] .
(13.2) 3 a= k .
~(~S Fiorri the defining eqriiitioii (13.ill (13.2: Example of a r.sided expotiential HPI. 1) w e obtain .ight.1)directly Figure 13.
11.I ’ The ztransform of t. Its ztransform i s 2 Z { E [ k j } = . Tlie houndary of thc .7) Figure 13. of t t i t ~ fundiorl X ( z ) from Example 13.5).1x1 > 1 (13. The a‘l’ransform I5giirc.2 For U = I the seyiicnct’ x[kj represents the unit step finictiori &[kj from ( 1 2.61 z .3: Region of convorgmrt.hc a lehitled exponriil id s e q i m i c e from Figure 13. which riiwiis all valiiw of z in the cornplcx i planc that licl witliiri a circle of radius a (see Figure 13.4: Exariiple of a.r[k= ak c[ k .6).4 . lefthidetl expoiicntinl scqiwricc This siim convcrgcs for /z/ [ U / . 13. n F Q is (13.318 I D .
a unique IVc are familiar wiih this sit uation from the l ~ p l a c c transform.2. This shooltf not lead t o the misintcrprelatiori that we are onlv dcalirig wit11 rnii1. < 1) b ~ l o n g a decaying exltoncntial stqucnce arid vdiies out.3 ~ s ~ i t ~ t of arid 4.Y region of tmivrrgencc is cirrular licrc too.idc the to unit circlc ( T 2 1) belong t o c? growiiig exponelitis1 seqiieiice. In ccrmptirison to Examplc 13. since d s o for k < 0. Tht.ai.5). In Figure 13. Values of z = I’c~‘‘ witlirri ihc unit circle (/.4). 1. wc irotice that the ztransform of thr riglit sided cxponential sequence (13A) arid the leftsided cxponmtiaI sequence (13. All other valiieq cm thc unit circle repr rit complex expotieni in1 oscillations ctf frcqiicncy 6 1 with ?r < i ic T . .J ) k .3 the region of corivergencc. and 13. ti’ignrc 13. In Exainp l ~ s4.1. This eiriph M’ithout it.6 the exponential srries z h are each oiily slrom7n for k 2 0.1 arid 4.lie w i t circle correspond to thc cqtonential serim eJc’kwitlicorrstant amplitude: z = 1 lei& to a series with constant values because eJok = l k =.1 is the highest repiesentable frequency. Definition and Examples  319 Figure 13.z k f 0.3. c‘omplcx coiijngate values oi 3 l a r t distingms2ied by the tlirectioii of rotation..13. valucs z = eJb’ on t. and its iadius is equal to tlie mngiiitiide of the pole. cari iriterprct the individual points of the :plane in a similar way t o the splane in Chapter 3.1.2 we corisiclcred lcftsided mid rightsidtvi continuoustime signals tlint arc’ likewise orily ( ~ s t i ~ ~ ~ iby ithe region ~ corivergence (see Fignres 4.7) h v e the sarnt’ form and only thc regions of convergerice axe diffeierrt (Figiws 13.w F ) x k = (.uteral scquences. whilc z = . ne Wi. 1wc.6 shon s the corresponding exponential secpen(*+’ zk for different values of z .5: Region of convergencc for the fiiriction X(z) from Kxnmplp 13.
rotation faster I maximum DC component rotation I .
.
i0 Close rdationships exist Ixhtwetw t lie :transform lmow. We will IX)W exitniii~p more precisely thcse relstiormshipr bet ween the iridividttal t rnnsforms. like the Lapinct.ritxs are closel~~ F . transform of ttii qucnce is irnmt~diately bhined.9) tlw Fourickr. [k]has been cieatrcl 1)y saiiipliirg a continuoiistiiiic signal x(t).trarrsfotorm of a sequence z[k1 i s only calculated for \dues ciiclc .1I ) .9).FoLrricr 1 rairsforrn F of R Istime signal (bee Chapter 9. i ~ 2 on the unit (13. transform ancl Iht. c t d other trrinsftorrns WP already The ztransform and tlir Fouriei t iansforni . x of a sc.10) (13. Bet t h e Laplace trwrisforin X ( s G{ r ( t ) } ant1 thc 25jfoririX ( z ) = Z{ there iritist likewise lie a close corinect ioxi. train (see Figurt.It E R .3). By cornpar iiig t lie o c instructions f o r cdmi ng the 3 tralisfoinl arid the Fouricr trilnsform vv rccognisc that (13. tlir valnrs of I [ A J c m be iriterprctcd as thc wciglit ings of individual irripii in iinpulsc. When the . 12.
13.3. Relationships to Othcr Trausforrnat,ions
323
13.:3. Kelat ionslh i11s t() Ot hrr l h1 rsforrrrntioiis
325
s plane
I
xaniple 13.5
’i’he ztransform of t h e discrete unit, step kunci,ioil
: Z { E [ k ] } = E 1 ’ / z /> I .. L
&[Xi (13.6) is
w% iiow arrive directly at, the Lapla,ce transfornl of the coirtinuoustimr signal
by setting 2
= P’.
The h i m e ho~clsfor practical app1icaLioii of ttrr ,z1 rarisform as for the Laplace i r ansform, w1iic.h wc liavc. alr y covered. Calculating 1 he zi r:~iic;forni a of qucnce I y maluating tJie w r r t h r r formula (13.1) only leads to (13.3) for queiices z[k;(compare Example 13.1).For otticr cases, wc try to refei t o these siniple sciirs by iising sonic gcnclral rules. Thvse grneral rule5 will be summarised by a series of theorerns that h a w n lot of similarities aith the thwrems
13.4. Tlreoreins of tlie :'l'ransform
327
of the Laplace trarrsforrn (see Cliaptrr 4 7 ) . T h a t is no coincideticcl. ab we saw in Section 13.3.2 that the ansforni of a series can be thotiglit of as : ( w i t inuoustimcsignal con g pmely of delta impulses. It slioulrl be many properties ol: tlie 1,aplac.e transform also apply to tlic xtransform.
The most impor tarit tlteorrrri:, of the ;tuansforrr~arc snnimarised in Table 13.1. They cair i)ISo be sliowri without rcferrirrg to the Laplace transform by iiiscrting them irrto (13 1). In contrast to tlic tlieoiems 01 the Laplace transform. the intlc pcnclei~t variable i s deliiied orrly for iiitegei valuc~.In thc square brackets [ 1, no rc pclrmit tecl. and corrcspondirigly. for the shift thcorcm, only in shifts /t E Z &reallowed. As for the similarity tlieorczn (4.241, for samplt~tl signals t i w orily scaling of the time axis that i:, pe'rinitted i s cx = 1, so that bccomes the time reversal theorcm for the ctraiisforni. Reversing tlie index of a series of values t m i be dorre simply by w d i i i g thcin backwards.
Tirnedorrraiii
zdomain
x X ( x ) 1 b Y ( z ) R W > RC>C{.cJ? ROC(y)
Shift,
Mult iplicatiorr by k:
Time :eversal
:328
13. 'L'hv aTransform
n
ti
xarnple 13.7
TIie coefficients can bc rcad clirelctly (coiiipart3 Example 13.1):
3
I; = 2
k = I1 k=l x 2 ot lierwiw
.r[k)=. 2'(X(z>}=
0
.
( 13.3 7)
2
3 2  1

 *
IT
3
4
k
iii igrirc 13.8: Series alk] ohtaiiircl from X ( z ) i i t E x a m p l ~ 13.7
(13.2%)
has two iiiimcwtor zeros (poles) outside the oiigiri of the cornplrx plarw: zL 1 2.5: + 1 : gives z p l = 5 , uxp2  2 As the rcgiori of corrvrrgenct. lies outszdr a 0 circlc arountl the origin, we exp riglitmirtl series r [ k ]= Z ' { X ( possiblr to esprc'ss X(zj as the f two terms (like (13.5j). m ' can y term an (~xponc~iltirzl oscillation. Diiect partial fraction expansion does nut lead t o the goal a \ it prodiiccs ;t b ~ i i ^ofi the form ~
(13.30)
to cictrmiinr the exponeIit,ial teririi. Hcre it helps to expmd X ( :/ )z ir'ractions, instcad of X (2).
into partial
(13.31)
Determining t h~ partial fractioii cocfficienls. for exarnplr. by cqiiating coefficients. I r ~ d s Ro = 1, 6'1 = 2 a d B2 = 1. Multiplying b) t and trrmb~icrin tr) iriversion aided by Exainplr 13.2 yields the series (13.32)
( 13.33 )
3
e
2
Figure 13.9: Polezero diagt mii for [ 13.25)
Figitrr 1Y 10: l'olezero tiiagrarri for (13.28)
ris
TT
2
3n &
4
1.4 =r: 0.6 1.1 x 2.2 0.7  Fz 0.15 1.5 x 3
~
hiit they do inakc ~1 h e a r roni rihut ion towards tlrc phase.11 with the exact t for comparison.28) gives A. circle that a pcrle 01 x r o lies.ho polczero d iagr a m Substituting z = 1 into (13.gni( ude of the Four ic. Zeros lying tiirertly or1 the unit cirrle forcc7 the magnitude o Fourier tiensforrn at this f r ~ q w r i c y zero. Poles 01 zeros at the origin of the zplane hav influence oil h ma.28) and the estimate from t.ier transform. has on thc h"mr.1 trarisforin. The estimates are show11 in Figwe 13.= 1. the greater PB it. .F'igiiri. and to [ d e s lying w a r to tlw niiit circle incrraso the uiiit. 1311: Enact frequrncy rrsponsr to (1 3.
[k]) and sketch thc given sequences arid rrmgriiturlcs of the spectra. Xotc: bee Exaniples 13.7.[k]. iisirig the rleiinirig equation (13. xercise 13.~.2 and 13. responses zl.4 a ) Fiud thc xtransbrrns of the lowpass inipulsc.C.(. Exercises i~tid give  335  ~ the region of convcrgcncc for cach.13. 0 I ot>herwisP p=l 2 b) Give the sprtra . .3.I ) . Prow the followiiig theoreins of the ctransform: 1. xercise 13.
the inverse 2transfcxiiis of thc following fiiwtiouh: . xercisc: 1 Dc.H:.(z) = 7 c 1.1enniiic.
Excrcises 397 .L3.7.
This Page Intentionally Left Blank .
5. In fa. This is in a.ct. \I'P then fall back on Definitions 3 t o 6 fvorn cair Chapter 1.1) .Nom7 that we have described tliscre!tetirne sign in the frequency domain.1 yields the lollowing: A discrete syst.etr.s I h p properties of discreLe syst.rire y s t ~ m s . The properties of the discrete spterri S will be restricted so that it is an LTIsystcrir.liein inore quickly by buildiug the com?sponding properties to continuous systems. (14. we ca.rneiitalrelastionships.rel. s but. Figure 14.isc.They have already been 1xiefl. signals.y mentioned in Chapter 1. The system S pro( the input scrics ~ [ k ] produces and the output smies y[kJ. we ca. were left in favour of continuoirs sy . I ? ity a The niost geneial form of discrete system with o~ic input and on(>output is shown in Figure 14.u deal with t. as thrir generd forinulation applips to both continuous aiicl discrete systems. and S ( h 1 [k]+ B~:![k:j}A S ( ~ i [ k + B = j} S { ~ 2 [ k i } ..1. as we want to avoid technical realisathis of discrete arid coritinunns systmns and deal oiily with t2ie funtla. W e call these d. I: Discrete LTTSystem Specialising the superposition principle and timeinvariance to the system c k scribed by Figiirc 14.n Cled with systems which have discretelime input and output.rd. signals z ~ [ k ] :r:z[k].erxis have strong pa.2.crn S is h e w .ll(:lswith continuous systJemx. . if the sliperposition priiiciple holds for the response to any two input.ccorrlance with our general goal.tems.
In any case. .K inlist give aii integer s index of the series ~ [ k ] . 4 aiid R can be any complex constants.2 ] .7 ~ [ k .l .. gJ[O]. p &N t11. a shift of the time index C by n values ocxiirs in place of the nth derivatives.1 Difference Equations As 1irietLI differential equations with. for example.6) is a discrete LTIsystem. the systcm S is called ship?n.3. linear diffcrcnce equations with constant coefffcieiits cherarterim: discrete LTIsyslenis.ciarzcxnt. h.pjkl] + y[k’2] 3 3 = = . V K E (14. DiscreteTime LT1Systcms As in (1. Exarnples for such difference equations arc 1 .K ] = S{lr[C.340  14. that caii iii general be formed from I independent linear V wiiiponents.y[1].the difference eyuatioii (14..~ ~ t l e r linear tiiffermce equation with constant coetficients a. .6) has XI infiiilte number of solutions y[k].r[k] 1 ~ [ k ]~ [ k . For a given input signal cr[k].yik] + y[k1] 2 1 1 ~ [ k ].3) holtfs. is In contrat to the rorresponding form (2. .2) t he relationsliip g [ k . y[2j. const ant coefficients rorrwpond to miitinuous LTlsystenis. h1 imwmbiguous sctltition can be obt2tined with N conditions. a tlie digerence X* .3). 14. a discrete position variable.K j ) . Using the superposition principle (14. given i by initial vdiies instead rimtives. for example.3) of differential rquatioii.2) and the condition for tineinvariance { 14.5) The general form of itli N t l i .. the shifl is oiily tlcfinied for integer valuw (if PG. + ] 2 (34. If the index k is in ad of time. A discrete system S is famcznvarznnl if for the rtv3ctioli y[k] to any input signal ~ [ k ] y[k] = S { z [ k ] } (14. it can be confirmed that a systcAm described by the difference equation (14.4) (14.3).
y[k] rrJpcresents the solution of the diflererice etlitation iiiider the initial conditions. 2 .7) cl. BY the right siclc of (14.4. we obtain the expression (14. which corresponds to Figure 7.1.7) can be calculated quickly arid very accurately using a digital cwmpiiter. U[N I]. The exkraal &lid internal tclrms C i i D be calciilated in a similar way u5ing the ztransform. The fundamental priiiciple is agairi splitting thc solution into an external and internal part (14. .2. are entered into the rightlmid side.3.14. The effect of the combined terrnh is shown in Figurr 14..1 Numerical Solution Ky rearranging the diffcrciice equation (14.oan be interpreted as superimposing eigensequences of diwrete LTTsystems.1 for corrtiriuons initial value problems.22 AnaXytical Solution Therc is also an analytical solution which is possible in mut11the sanie way as with t lie linear differential c.y[k 21. This mrthod has practical importaricr.5 we rnentioned that the inverse ztransform. just as the Laplace transform was used with coiitinuoucl systems (we Chapter 7).yext[k] describes the response to an input signal if the sy icst aiid ylllt[k]describes the honlogenoiis solution of the iuitial colidition problern without an input signal. P ~i~~~sequen~es me In Chapter 13.6).2 Solving Linear Difference Equations The are two possible methods for solving the linear difference equation (14. yjk .1and 13. 3 .3. .quation. . The broperties of . . . known. Characteristic Seyueiices arid System Functions of Discrete LTISysternd341 14. the values already with the initial d u e s g[O].. A delay by n can hc iniplciiicnted by shifting the location of the values in the memory of the computer. For k = 1. For k > 1. ~ 14.8) Here. the that can be numerically eval calculation of the initial con ns starts with the known inpul signal z [ k : a i d .6): rrurrierical aiitl aria1ytic.. . .2. begirirririg with X = 1 . 14.3.il.11.
4)= s { z k .9) F'roin the conditions of timeirivarittnce a i d linearity: y[k  /€I = S{r[X ." s ( Z ~ } = :"y[k]. (14. (14. (14.2.3) The p r o d that eigenseries of discrete LTIsystems are exponential series of the form c[k]= z k will be carried out exactly as for tlie eigenfniict. z €42 (14.2). id y[k] = S { 2 } .342 13. An eigenseries at tJie iiiput of a discrete LTXsystem causes a response at the output that corresponds t o the input serieb with a constarit factor (see Figure 14.1) and only require that.10) if y[k] is These difference equations are only fulfilled at thta same time for any a weiglitetl cxporiential series y[k] = x z h . We want.ions in Chapter 3.1) arid timeirivariant (14.12) . (14.2. We consider a genrral LTII Figure 14.11) Equating x [ k ]= z k and y[k] = Xzh yields that tveiy exponential series t[kj Z X k . to f n thc corresponding oikpiit signal. K.3). IJiscreteTime LTISystmis initial states I oneffect output I Figure 14.3: System S is excited by the eigcmcries c [ k ] system (Fignre 14.2: Combination of the external and intcrnal tcrnis to forin tlic solrrtiori of a differential equation eigeiisequences ale completely analoguotis to those of cigenfiinctions of continuous LTIsystems in accordaiice with Definition 10 in Chapter 3.2. The iiiput signal should be an exponerrtial series a [ k ] = z k . it is linear (14." } z .
17) . the region of cwivergence of H ( z ) is oftcm riot exglicilly given. The (*o1 f responding eigenvaliies X depend in general on z .4: z k arc eigenseries of discrcte LTJsystrems. which is given here as tlie inverse transform of the input sigrial n:[k]= y 2n3 and the output signal dz X(z)zL (14. Finally. Exciting t he syst em with a complex exponential sequence outside the region of convergence does iiot lcad to an output series with finite amplitude. As with contiiiuoiis systems we call X = H ( z ) the s y s f m f u n c f m n n of R discretc LTJsystem ( function for contiiiuous s in fiinction for discrete hystenis Ff(s)has a region of roiivergence which contains only part of the roinplex plane. it should also be nientioned t h a t a unilateral exporieribial series z [ k ]= Zk: .[k] y[kj = S { z [ k / ) = dz z (14.r[k]. Cliararteristic Sequences arid System Functioiis of Discrete LTISysternd343 Figure 14. (I 4. H ( z ) is the eigenvalue is an eigenseries o ari LTIsysstem.4.14) (14. l ' h e relationship between a genmal discrete input xeq~~erire[ k ] and an exponential sequence is represented by the z ztransform. 11s with continuous systeins.16) Y ( 2 ) = H ( z ) X ( z ).15) of an LTIsystmi.13) not an eigensecluence of a disc*ret!esystem The system function H'jz) ran describe the system response to all input sigiials . € [ k ] is (1 4.14.anti riot just eigensequences e [ k ] = z k . Thc output sequence y[k] is obtained as the system response to .
1).344 14. Example 14.11.6: System €imct.11) = z .' X ( z ) .1 A discrete delay circuit delays the input signal.18) = 1 yields Y ( z j = Z(y[XJ)= Z{. (14. N e will demonstrate the prowdure 7 with a few examples. nctions from rence Equations A systclrn fiinrtion can easily bc found from a difference equation using thc shift theorem of the ztiarisforni (see Table 13.5: System function H ( z ) of a.6) H(x)= 2 .I . discrete Ill'lbystern The system finirtion H ( z ) (with it®ion of convcrgmcc) is i completc dcl scription of the input output behaviour of a discrete LTIsystem. It makes it possible to give the output sequcncc y[k] for any given input sequence xc[k]. Its 'difference cquation is y[k] = IC[k .19) 4 comparison with (14.C[k . Figure 14. Using thc ztransform and its shift tfieorcrn (Table 13. (14. . DiscreteTime EI'ISvstenis Figure 14.ionof the discrete delay circiiit .1 ) with /.17) givcs the system func*tionof the discrete delay circuit (see Figurc 14.
I] .i .2 = . Wjth the two exarnples and with (14.25) can be found.14. N N Y ( x ).21) Froin Ihe difference equations.2X ( Z ) .T[k 2 1  2).4.1 + y[k .24) und (14.20) Using the ztransform with its shift theorein Y ( z ).6) can be obtained in the same way as the last cxample. Z 2 .). 1 + xjk .[k] 1 1 5 3 1 . ( 14.24) (14.k(z)zZ 3 = X(*>+ x ( z ) x .6.. n*Lzn= X ( z ) I (14.2 We will firid the system fiuiction of difference equ' t 1on CL ' y[k] . t discrete L'L'Isystmi d e % ~ i h c d thc I)y 1 1 g[k .25) we can make soin$'statements with which IW are already familiar in priiiciple from dealing with contiiiiioiis systems mid the Lapbce transform: .ric eqiiatioii is formed so that X ( z ) m d Y ( )can be factorized z Sorting the terms yields tlir system func:tioii The system function of a gmcrnl LTXsystem of Nthdegree described by llie difference eqtintion ( I 4.Y ( * ) / 5 I + . (14.rristic Seqnences anti Systerri Functions of Discrete LTISyslrrriB45 Example 14. The ztransform arid shift t licorcrn yield from (14. Charact. a n a1gebr.
its region of convergence can only he given if additiorial properties like causality or stability are known. Comparison wi(.25) confirms that the block diagrani in Figure 14.7 realises a discrete LTIsystem with the difference equation (14.2. The only differences are the its in place of integrators. it is also striking that both vertical series of delays run in pariillel and ran bc replaced by a single series.. The system function for a difference equation of firrite ordcr is a rational frncl2on fuwtion a d can be describrd by poles ancl zeros. systcm kunctiori H ( 5 ) = sl (integrator) corresponds to a block with (lie system function H ( z ) = z’ (delay circuit). Discre te’l’iiiieLTISystems The ztransfoformconverts a linear cliffercncc equation with constarit coeEcierits into an algebraic equation.6). Like mrttinuous systems.1.5.9 again corresponds to the structure 01 ct form I11 (see direct form 11 for the continuous system sho.he region of corivergcmce lics outside a circle about tlie origin. Q Wheri the system funct. there is tl delay in clistrct In the heqiiencydomain it niearis that a hlock with the. shown in Figure 14.346 0 14. form I of a discrete lJTIsystem as in Figure 14.6. through the most tlistarit pole.vri in Fi ns.5) can be coristrwted in a sirniliar way for . For causal systems t.21. The result in Figure 14.ioii I_i(z ) is derived from the difference equatiori. The block diagram in Figure 11. but tho result is the sairie. The frtxyency response of the systerri can be estimated easily using the methods described in Chapter 13.h the system function (14. Figure 2.7 corresponds exactly to iagrani for coritirmous syterns iri Figure 2.8. ‘l’ho order in wliich they axe connectetl can be reversed (coriipare ure 2.7 t a n be analysed as two LTIsysttms con~ie in series. 4. h s in Chapter 2. cfsrr ysterris can be represented by block tliugranw The parallels betmeea the two kirids of system itre particularly noticetible here as instead of an integrator in continuous syStcms. The algebraic equation can be solved or investigated niorc easily than llie difference equation.
rm 1 of a tfkcrete LTJsyst.(14. Block Diagrams arid StateSpace 347 1 Figure 14.26) (14. a statespacc representation of the form (14..5. Using thc ztransform on the statespace d ription (14.27) vve obtaiIi: (14.27) it is easy to charzractcrise systems with multiple inputs niid outputs (Figure 14. As state vector ~ [ kthe vdnes stored in the delay circuits are chosen.10).11.Xi) the s y r t t m ecpation.With the form (14.26).placetlornain cxwpt that the complex variable s has in this case .28) (14.rm e s c r ~ p t i Q ~ Discrete for If a block diagram of an LT1system is given.29) Thesc cquations are identical to the stat.27) rail be formalated. which ckscribes changes in the intcriid states depending on the current state zik]and the input series x [ k ] .espacx! description of U continuoustine systern in the l. W e ] c d l (14.7: Direct ii.i.26) arid (14.
28).348 14. Figine 14.27) can be extcnrled. (14. block diagram arid statespace desciiptioiis (Chaptei 2.27). iri a suitable initial statcl ~[Xoj that tlit give11 so initial coxidi t h i s ale fulfilled. the statespace description (14.29).2G) and (11.1 2. 6 ) . eqiiivalent statespace rc~prcscni ations (Chaptc. ancl (14.space description is .d for contiiiuous systems directly oil disc. includes the relationships I. This can be achieved using a superiniposed impulse> zoSjk .ku]. We rim thcrclfolc also use mariy of the> techniques dwc4)pc. (14. tvlncre the system is offset to a suitable time ki.5) a i d controllability and observability (Chapter 2 .ions. Bc crtrefiil n o t to roiifuse tlic state vector in the timcdomain ~ [ kwibh the frequency variable 2 of the ] zt rallsfoKrn! In the case of initial condition problems de ibed by difference equat.26).etwec>n ~iff~rence/di~erenti the a1 c on.8: Forrning direct form I1 h r n direct form 1 hem icylaced by tlir rvmplex vaxiable z. DiscreteTime LTISystems Figure 24. with which the statc. I’hi:.4).10 shows t Ire block tliagrarii of the st atcspace description for a discrete LTIsystem with state qiations (14. systems.
. y[N 21.66).61).11. Thc procedure will be demonstratcd with an example. rlisrrete initial vdne problems can be solved in tlic usual way. y[.62). y[O]. .tcrn tion is z Z ( Z ) = AZ(z) f BX(2) (14.9: Direct form I of a discrete UT'Isystem T Figure 14. Block Diagrams and StateSpace 349 t Figure 1a. Starting from (7. . + .(7. Thc initial state eo can be riisily reached if thc initial conditions are given: y[l]. (7. which corresponds exactly lo the initial value problem foi continuous systems in the Laplacedomain (7.64) .5. .31) YfZ) = GZ(2) + D X ( z ) .10: Block diagram of the btatespace dcscript [on for a discrete LTIdyst.14.
350 14. z2[1] = 4y[l] I IE.s irnportancr can be expressed in three significant properties: .(7. This can he confimed by insertion iril o the givw diffeience equation.64) . D=1.espace description: . in addition to thc system function. and the initial values y[0] ] and y(Z].1I is described by the followiilg matrices of the stat.A=[1 : 0 B = [ i ] . C=[O f ] . 4gnal also starts a t k = 1.66)): = Y ( z ) G ( z ) z ~ . can be wad froin Figure 14.A)' = 1 = . thc calculated uiitput.AI 22 1 4 1 z].lroin that we obt>ain yt4 = + 1  (Y[OI z4 ~ ~ 1 ) (14.11 21/11 = y[O]. using the formula (cornpare (7. W e calculate the response of the system for ~ [ k= 0. which caiiseb the same output signal. In Clinpter 8 we introduced the iiripiilse rcsponse c2s a second irriportant characterist ic for contimious systems. Ii. and delivers the expected value g[11. 1 The conriection between the given initial values y[O] arid yjl] mid the statc vector zo = z[/]. the output signal is foiind by inverse ztransformation As we bid taken the initial state at XiJ = 1 . G ( z ) C(zE .32) 4 and splitbirig into partial fractions Finally. DiscreteTime LTISysteriis The secondorder discrete system in Figure 13.
2.1.6.c.inedin the timedomain by convolution of the input sigiial and the impulse response. . the selective property. It shows a discrete system with input signal . ‘l’hc impi~lse response is obtained by inverse transforlning the system function.t. In one poivrt the description of discrete syskrns becorncs a.2.3 The impulse response of a system is its reaction to a pulseshaped input I. for a discrete system.1 ~ a ~ c ~Q€~ a System n the ~ i ~ iscrete ~ o ~ i v Q l ~ t ~ ~ n To introduce the impulse response of a discrete system wc consider Figiire 14. which is a sequence of zeros and a single one.a. The output signal of a systern is obt.12. It i s the unit.s vihlues. impulse iritxoctimd in Chapterl2. 14. T h e response h [ k ] to a ctisc*r& unit impulse (see Chapter 12. discrete system.6. as the pulseshaped input needed to geiierate itri irnpulse respoiise is not an ordina. Nre can deal with these w. Discrete Corivalutioii and Tnipulsc~ Hrsporise 35 1 Figiire 14.14.scd on thc irnpulse respoiise can be transferred to that.1) will be called thc impulse response of a. . The delta irriplusr introdiiced in Chapter 8. 0 The dcriviition of blicse properties was not simple.ions.3 cxmnot be chara.te tlic impulse rrsponsc is now a very simple sec1uenc:e. in particular.11: Block diagram of t. tlic impulse response has the same fiintlamcnt.hr discrrtc systrm in Example 14.lues without the use of generalised fimctions.ry function. The price for the elega7nccof this system descripliori was generalising fiinctions to distributions.c[k] arid output signal pylk].al properties as for a contirmous system. lot simpler: the pulseskiaped input signal requirccl to genera.erised simply by it. only by its effects on other funct.
z [ ~ . ] h [ k .12: Discrete coxivolutiori can also be extended to any inpiit sequence z l k ] with more than three values.]6[k K ] for I$.he input sequence x [ k ]in Figure 14.hr. and then these sequences can be interpreted as a. Eacli impulse is .mine t.. so ylk] = ~ . = 0.. 2 . c .unit impulse shifted by IZ = 0. for t.l i ] IZ = 0.].352 14. Because Llie system i s linear./.iveiglst.o I L ' \ K ] ~ ~. We can split it into t. The sumsnation is performed in the general case from OG t o 00.. The resulting cxpressioii for y[k] is tlic same as for convolution if t l ~ c integral is ieplaced hy a suiiimatioii sign We therefore refex to this as discrete convolutzon. which will be denoted with *: (14.ee individual sequences with one nonzero valiie in each.1.he responses to the individual sequences can be added to give t k e response to the 2 coniplete input signal 4 k ] .c I$.cdby the value of' the firnction :~. :r:[t. 1 . At the output of the so LTIsystem each of the shifted aiid weighted inipulses c a i ~ s e likewise shifted and a weighted impulse response.12 with three nonzero values. DiscreteTime IXlSystems The impulse respoiise niakes it possible to calculate the system resporise in the timedomain for discrete LT1systenis. These considerations x[2] h[k21 0 1 2 3 Figure 14.3) . As an exainplc we exa.2.2.[k].1.
14.tlisrretc c. on poles of H ( z ) it they axr at the same time zerob of X ( z ) (aiid rice verba).n ) .ion ) I = k . we start with the discrrte convolution in (14. t h e relationship between the impulse response h [ k ]and the system fuiiction H ( z ) reiriains uidefinrd.0 0 z[Li]h[k:4 z .12. ) and H ( 2 ) .ergeiice ctf Y ( : ) (iin thcrcfore br gipater than the irker The secoiitl $t. The first statenient identifies the conwolutzon th~oierriof tlie ctransform. E'rom this ~ v e rrad: 0 tJsing zlr~nsforrn.36) U ( : ) is at least tlie iritrIscctioii 1.stement gives Llrie relationship betweeil tlie irnpriise respoilbe :i. Both sLms dTP ind(ipciideiit of each other arid represent the . 0 Tlie 2transform d the impulse icsponse of a discrete s y s t m i i s the system function.6.transforms of t [k]arid h [ k ]respectively. (2) ROC{!/) _> ROC{z} n ROc(A}  (14.corr\.33) ancl use the ztraiislorm on imth sides: U(") = I. Y(L) also conveiges. tvliirh corresponds to t I i c convolution theoleni of the Laplace trclnhforiii.oiwoliition becomes a niti1til)licatiorr of the corrwponcling ztransforms.K 0. TO liiid it. T h t means lhat Y ( z ) definitely convelges whew both X ( 2 ) arid I I ( z ) also coiwerge.Ltion.ncJ system function: The legion of contergcwr cif (14. Iiowwcr. Tlie region of. =m y[k]+ = k=xK. Discrete Convolution and Tinpulse Response mu t at ivc. Iiitcrchengiug the slims and rearranging terms yield5 111 the scco~ld row the substitu(.  (14 24) On tlie riglitliancl side we have added in the exponent of s anol her L P I O in the form ot K . 353 Con~olution Theorem of the zTransform In tho dwriptioii of the system from Figure 14.etween the KOCs ot X i .k f ( f i .37) .K has beeii used.
Depcnding on the structure of the syStein. for rxmrple. ronsists of a direct path from the input to the output and a parallel yzth with a delay arid a mult$lication. From (14. DiscreteTiine LTlSystems As well as the s>steni fiwction is the impulse response a coniplete description of a discrete LTIsystem's input.37). % input signal: unit impulse % multiplier memory (initially empty) xold = 0 . Example 14. It.354 14.4 Fignre 14. for k=1:6.8. y = zeros(size(x)). 1 will consider two exarnples for discrcte lirstorc'lcr % b\i stcms. of the following commands: x b =[1000001. The response to an impulse therefore consists of txm valmcs arid is of fixritc length.it can have a finite or infinitr number of nouzero vzlurs.13 shows the structure of a system with system ftinrtion W ( x )= 1 t bz' . 14. = 0. impulse response ++ k Figure 14.out put behavionr. % output signal (not yet calculated) % loop f o r 6 steps . Thc system resporlsr to any iiiput seiics can be calculated using convolution with the irnpulsc rtxponse. ysterns with Impulse nite Length A significant distinguishing feature of discrete system is the length of the impulse response.13: Systmn with an impulse response of finite length h soitware rcalisation of this systerri in the programming laiigiiage h'LU'LA€3 could consist. which in Section 14.4 we assumed was already known. the Iegion of convcrgence ol the system function N ( t ) can also be fbniid.
y = zeros(size(x>).. Figlire 14.d = 0.1 7. k end % memory (initially empxy) % output signal (not yet calculated) % l o o p f o r 6 steps % current value of the output signal % delaying of the output signal IIere i8ki(> ciirrciit output sigiial is s a w d in the variable yold. % input signal: unit impulse % multiplier y0l.x 2 . impulse response t nL“’ = I 0 7. .so (Iial t h d e l a ~ ~ cmi be irriplerrieiitect.rl direct path and a fedback loop from the output to the inpiit with . () end % current value of the output signal % delaying of the input signal The delay i s formed by saving the current. inpiit value in the variable xold. [l yold = y C l . Discrete Coilvolution and Irrigulse Respoilhe 355 y k = x k t xold*b.(3. f o r k=i:6. () () xold = x k .a2’ I(. y k = x[kl t yold*a. nk f o r k 4 0 elsewhere Figure 14.8.t delay and a rniili.iplicatioii.14 shows the structure of a system with system fiiiiction H(2)= 1 .AR arc x a =I100000l. is rriatlr iip of . = 0.14: Systerri with an impulse response of rrrfinite length The c:orrcsponcling program steps in MAT1. The response to an impulse tlecays stcadily for 0 < 101 < 1.14. 1 . hut never quite reaches zero.6. ‘l’hc inipulsc response i tlrerefore s infinitely long.
? K=m . Draw 44 and h [ ~with I ] 3. In the usual programming languages. DiscreteTime LTISyst.ansform of the discrete writ step iir Chapter 12.4.38) c m be implcmeiitetl irnrrrcdiately as a computcr progr i>iYj. Miiltiplicatiou of 2 = [ ~with h [ k ] and summation of the product for all values ol I(: yields orw vdiic\ of y[k].ion of cliscretc cymvoliiiion iisirrg tllese 1 in sonw examples.I[ ~ . j ~ r /. We calculated the F*tr.dciiIat. The irrrpnlse response of tlie accurnulator is the discrete unit Figure 14. . 5.Finite I m p u l s ~ Response).2.356 14. and the irriier loop via the index IC. We will therefore dea1 with it.3: 1.cnis For a = 1 this system represents ari nc. th 4 eps (*ail also be carried ollt es.by k positions to the right: h[r.~[x.3.] + h [ k . ] 4. two FORloops are required to calculate (14. Systems with iiifinit.3aiid the corresponding ztransform in Example 13. Repeat steps 3 and 4 fc iise of the comniutativity of convolution. liowevw. wlierf the outer loop runs via tlie iiidex X.kj. The folIotviiig tcrnis will 1 ) i ~s t ~ i : i Systems with finite impulse responses are also called nonrccurszve or FIRsystems (FIR . processors with ii special archit ecd i i r e (digital signal procclssors).38).3. Tile c. To begin with we show a rnetJiod similar to thc convolutioii iritegral in Chapter 8.e impulse respoiwe are alsct called rem~r'szueor TIRsystems (IIR Infinite Iinynlse Response)& 14.c*imiulator that sunib all incoming input values. iii even greater detail Lhan tlie calculation of the convolution integral in Chapter 8.6.14 for a = 1).[pi] h[k ICI (14.] .38) as lie scalai protlnct of t W J vectors very khtx preparation illid testing of such programs it is vital to have convolution on paper.~ .4 Discrete C o ~ v o l ~ t i o r i practical irnportnnce of discrete convolution comes about because [lie expression i I y :/[x. which cmi carry n with K iri (14. Tliere are.4. Sliift h f .3.
(14. Discrete Convolutiou and ImpuLsr Kesponse 357 Example 14. For the mulliplicstion of r [ / x ] and h!X.6. the sequence hlk .15: lnput signal ~ [ k ] impulse response h [ k ] arid By rcvcrsirig axid shifting thc iinpidst.16 fo1 k = 2 and k = ti. a5 shown in Fig 14.K ] and X [ K ] do not overlap (sec Figure 14.E ] i s 01)tdned.39) to an iriput>5ignaI of the form z[kj= as a resull of the convolution 1 0 forO<k<KJ otherwise (14. 1 and the suiriniatioii with i n d w TC there are three cases to consider: for k < 0 h[k .41) ~ The impiilse response h [ k ]and the input sigrial ~ [ ka] r shown in Figure 14. 0 K k 0 K k Figwc 14.16 bottom) and tlir pxoduct is zero for all E .15.42) .6 We calculate the resporise of the rettrsive system from Figure 14 14 with the impulse refiporise h [ k ]= & [ k ] d 0 < n < 1 (14. * (14.14.6 . response. . so yik] 1 0 .40) ?&I = X[k] h [ k ].
It is summed over ’ 0 k < A aiid yields (14.18).+] begin to ov(~l. DiscreteTime LTISystems 0 K k=2 k=6 0 K 0 Ic Figure 14.pthe inpiit signal z [ k ]for k <0 0 for 0 5 For 0 5 k: < K .358 14. thc prodiict CG[YI.E ] (see Figure 14.I. The srrrriiriation yields For <: k a11 1%vdues of the irlput signal ~ [ koverlap with the reversed ] and shiftecl iriipldse response h[k .ip.17). h [k.]A[k.44) .16: Mirrorred inipulsc rc?spnnsrdoes not ovev1a.[.h] arid .K ] so IC 5 k is iionzero (SW Figure 14.
18: Overlap for K 5 k K The sum of the values of y[k] from all three romponerit regions is represented iii Figure 14. ’a I 0 . Discrete Convolutioii a i d Im~u1se Resrmnse 359 overlap Figure 14. U  k Example 14.19.14.17: For 0 5 k K K < A.6.7 .the two functions begin to overlap K Figuw 14.
20 s2mvs an example for A = 2 Az[k .8 Figure 14. Superiniposiiig the coriiporierit.20: The giver] series i s shifted by 3 arid scaled by a factor 2 If h[k]is only a short sequence (e. m d ko = 3 . whose siini over K = 0.  Example 14. convolution products h t are given by the cortvolution of x[k]wit. Convolution of r [ k ]with each p a r t i d sequence (left. 2.k o ] . 2 . K' = 0.ko] Figure 14. S p l i t h g hjk] into shifted slid scaled unit impulses k [ ~ ] b k [ 4. Discrete'L'imc LTISystaris EI Tile convolution of a series .o cnlculatte tbc convolut. 1.45) 1 Figure 14. (14. 1 .12 can be carried out t.21 and the short iinpiilsc iespoiise h [ L ] . impulses h { ~ ] S [ kK ] .) gives the partial proctucb (right) y&] = h [ K } 4 k.h the sliift.ion product h[k:* ~ [ k ] : 1.22 shows the use of these steps on the input signal z [ k : shown in F'igiire 14. scaled unit.2 yields the complete output signal y [ k ](bottornf.360 14. .6) . as ixnpufse responsc of a nonrecursive filter).r[k]with in a shifted and scaled variant of z j k ] : shifted and scalrd unit inipdse results ~ [ k ] * AB[k . the steps in Figure 14.g.etl.
B k Figure 14.k]) I__ = duration{a.22: (Mculating the convohztion producl o f 5 [ k ] with the series h. f From Figure 14. The duratioii of y[k] cart be ohtaincd from the diiratioris of ~ [ karid h [ k ]with ] duratjion{y. For the last value of q [ k ] the equivalent is tiue.% n[k] = % .46) . 1 ( 1 4.14.p . Discrete Conivo1iiti)n arid Impulse Resporise 36 1 b k Figure 14.I .21 : Input signal :c[k]and impulse response h [ k ] .23 it can be taken d2iat t h e first nonzero value of the result ?j[k] lies at the index whir+ corresponds tJothe sum of the indices of the first valnrs of x [ k ] and h [ k ] .[k] C:oiivolntion of two signals o finite duration iesiilts in another finite higrial.jk]} + tiuration{h[kj} .6.i k *" x = k .
1 Cliwk the liticxity a i d timeinvariance of thc following systeins y[k] = S{z [ k ] ): a) y[k] = n s [ k ] I)) y[k] = .r[k . ke2 / k 1 xercises Exercise 14.r[k] d) ~ [ k ] uhxz[k] = e ) y[k] z &[k] x [k .362 14.1 1 k /k=O .5j c) y[k] = a f .DiscreteTime LTISystems t.
51: + u[k .3 A s> stem is given by tlir diffvrerwe eqnatioil y[kj = m[k] .2y[k .yjk . .r[2 k ] c) Downsampler followed by a ~ipsamplcr: Investigale espcially a delay by ~ i i stcp m t l by two stcps. arid b) x [ k ]= E ~ A numc~icallyo o k ] fCr Exercise 14. a ) Downsampler by a factor of 2: y[ki = . Exercises 363 Exercise 14.4 Solve t h e following di difference equation.1 .[XI Compute nurnc~ically outpiit signal of the following system for k E [O.Give the tfint:k] arid the initial corrdition y[Oll if the out piit of the dela~clement has the initial state z/O]= 2 c ~ r i t €~ [ k=]~ [ k. e Exercise 14.2 Are the following systems shiftirivarimt? Justify your answer.3.t[k]. e initial ~aliie pioblcm Irumerically for k ~j[k] c :O.21 . iriilial cwnditions: 2 1..14. 1 Determine the reaction to a ) z [ k ]= 6[k.5 y.21 = r [ k ] = (l)kE[k] y[O] = O g[1] = 5 excit ation: Exercise 14. ] .7..
he init in1 s t a t c iniiquc. X ( z ) .Stwt witohY.9 a) €~alciilate impulse response h l [ k ]of the discrebe system the Are we dealing with an FIR or i3n LIR system (jiistification)'? he approximated with a u FIR system H~(z). Determine for each cabe the initial statm of the delay elements in order tn fulfil1 t.7 Give the direct f o r m I1 and 111 for tlie system from Exercise 14.. For which forni is t. 1 I 4 () 2 1 k ~ [ kwith the invcrscl ztransfvrin ] xercise 14. ( z ) = AZ where zP is tlie pole of N ( z ) .he initial vonditions.11) Srdve the initial value problern in Exercisc 14.2 = ~ j k . l Exercise 14. 1 . xercise 14.3. Exercise 14.364 14.y[k 2 Calciilate the response to . 2 X [ k ] +0.5 and compare the cocfficicnts with a).5:c[k .5y[k . zp Thprl detwniine A in the timedoniaiii using t.r[k]= 1 . S C P ChapLpter 7.'! Ilrrive B difference equation from tht. A system is given by the difference equation y[k] .") &id produce i block diagram rcdisation.l] 1 U[ol = 3 r/k] = &[k] a) Cdculate ?j+xxt0 I&] 0 Ybxt() = H ( z ) ..5 analyticallyusing the ZtraIISfOrKK y[k] .3.5 into direct form i. Note. z ~ ~ b) Calviilate glllt[ k ] . IT and 111.l ] f g[k .4. Find Hz(.0. 2  . Determine the initial s t i i t e for each so that the iiiitinl conditions are frrlfilled. impulse rcThe spoiises of h l [ k ] and h a [ & ] should agree for k 5 5. DiscreteTime LTISys~enm Exercise 1433 Transforrri the block diagram from Exercise 14. block diagram from exercise 14.lie iiiitial conditions..1 =.
14.13 Sketch the corivohitioii protlnct y [b]= 2 [k]* h jk]: k k k . Exercirse 14.12 Solve the initial value problem from Exercise 14.4.12 Using the conivolution s l i m (14.21 with sketches. Exercises c ) Give 365 y[k]. the resuh with Exercise 14. .14.8S[k t I]. k k Exercise 14. 1)) ~ [ k= O . Compxe the results with Exercise 14.5. Exercise 14. Deterrriiiie its impulse rcsponsc using the idea of discrete convolution. calculate c[k]= a [ k ]* b [ k ] .10. K k ~ [ X ] ] .4 analytically with the 2t r a i ~ s f Q rCalculate the external and internal terms as in Excrcisc 14.for a ) ~ [ k ]~ [ k ]  1 3 : b[b]= S [ k ] 2h[k . ExePGiSo b[b]= 6[k + 21 + 0. Verify ~.7.1 1 + ~ 6[k .33).14 An input arid output sigilal for a FIR system are givcii.
DiscreteTime LTlSystems k .366 14.
Y Our examination oi systems hiis so far given little thought to akietlier or not they tail really be iiiiplement8ed.Altliougli in some of the examples we started fiorn aii irnpleruentation (e.. ‘I’be exavniiiatioii of cmsality is closely related for continuous and discrrltetime systems. theri oritpul). for exarriple. These new tools caii also br iisc. cariiiot happen before it has I .fu1 when the time and frPyuenc.1. First of all w e will be consitlcring the gerieral coxrtiiiuous systcrri froim Figure 15. we will iiot be coiiceiried with impleirientations using a specific tedi1rology. T h e socallccl analytical signal folloms Gom this u\e of the duality piiiicipke. In the lariguage of sigiials and systems this inearis rnore than just a logical corinrction ( I / iiiput.ycloniainsare exchanged. heginning with systwis in gelieriil and specialising at first into linear systems and then firially LTI sl sterns. arid fvuricl system descriptions lihe the impulse respxise or transfer fimction. bctwew the input and out piit of a qystcm. a11 electrical netwou k) . we will ht. looking for general criterid t h a t have to be fulfilled for impleriiriilation to be possiblr These criteria can be simplified iiito two concepts: causality a d stability. In this chapter me will be learning a siiiiple t imedoriiaiii characterii. so WP will be dealirig with hot h kinds of system in parallel. We will be introdiicing the coiicept of causality in step&. we have not l e t irivestigatetl the revel implementation of a givrii systrin function JYr will correct this in this cliap anti the next in accordance with our stated goals at the s t x t of the book. In this chapter w e will deal with quest ioiis connected t o the causality of it s y s t m The problerri of sy ai slahility will be dealt with iii the next chaptcr. thm output).atioii using the iinpulse iesponse Its extension Lo tlir frequencvdomaia their leads t o the idea of the Xfilbert transform. it also nicaris that an event. A causal relationship between t h e . Instead. Causality in general signifies a causal relationship between two or niozc processes. r c ~ icaused by another (first input. which has no special properties that we t a n use.g.
r ~ ( t and tor all ) time points t o .1: Definition of cauriality for continuous systems must he tulfillecl for all possiblr pairs of input sigrials 2 1 ( t ) and .y2(t) of the output signals.3) A linear bystriri is causal if an input signal r ( t )that is zero until time Eo causes an output signal y(t).1) Figure 15. be true for all tirnc points t o .tor k < ko.1) corresponds to the difference y ( t ) = yl(t) .te systems 1 With lincrtr systems.conditions.4) . (15.. of caiisdity for discrc. the condition (15. the difference r ( t ) = r i ( t ). that is also ZPIO nritil tirrrc to. then they will give rise to the same outpiit signals u l ( t ) and y2(C): rift) = ~ ( 1 )for t < t o =$ . j15.rZ(t) are the same up t o time t o . Causalitv arid the Hilbert ‘fiansform input arid output signal exists.2 holds the corresponding coridit ion for the discrete point in tinic ko :rl[k]=xz[k] for k < ko =+ yl[k] = vzjk] . The coridition for causality i s thus more simply given by r ( f )= 0 for t < to ===+ y ( t ) = 0 for t < t o (15. For discrete systems. Figure 1 5.2) discrete system Figure 16. (15. yl(f) = y 2 ( t ) for t <to.2: Definition.368 15. The correspoiiding causality vonditiori for linear discr systems is siinplg ~ [ k= 0 for k ] < rt. This condition rnust. AS neither linearicy ~ i o r theinvariame are prc.1) We could also say that if two inpiit signals 2 1 ( 1 ) arid ra(t)are not different up to a time poiiit t o . the output signals may iiot differ until this time point is reached. if two input signals sl(t)arid .rd(t) of the inprit signds from (15.0 y[k] = 0 for k < k o .
3) and (15. Causal Systems 369 LTIsystems w e t irneinvariant as weU as being linear. for cxampk.it) = 0 tor f < 0 r [ k ]= (I for k < 0 * ===+ g(t) = 0 for t < 0 i 0.0 for t < 0 is then fulfilled when tlie impulse resporrse h(t) disappears for t < 0: h ( / )= 0 for t <0.Lndity ot LTIsystems car1 also he expressed con< words: + An LTIsystem is ca1lsiil if its impulse responsc is a right. Tlic condition for c. fo = 0 or ko = 0.7 ) d r (15. For ront.2 is zero.5) (15. a i d so thr mnditionr (15.8) we can cxpress the coaditions (15.2 inclieate the beginning of the signals to be .o l l o from (15.30 and 14 23). 33 h( z ) x ( t . In Lhe case iriwstigated liere t .11) Figure3 8.33). ( 15.15.. obtairiirig for corltimlous and discrete systems: . disappears for negative time.9) 'l'bc r n u s a l i t y condition y ( t ) .16 to 14.l slid f. m m d u t e d (sce Figures 8. = tal = 1.6) significantly more concisely using the iniptilse respoiase. (15.l t .8) that ~~~ h[k] = 0 for k < 0 . = t. ~ where t.18 hhow calculations of convolution w i t h tlie i a i p i t l s ~ response o cauhai systems (contiriuous ail(€ f discrete). We have already learnt from these examples that the eorivolution result is nonzexo on\y for times later than t . y[k] = 0 for k (15.7) with z ( t ) = 0 for t < 0 that (1 5.6) With help from t h e coilvolution relationships (8.ixiuuus sy ins it follows from (15.18 arid 14.K ] (15. .sitled fiiriction and it.39) arid (14. .10) n tlie same way.4) only need to lte fornmlated for one point in time.5) arid (15.7) g[k] = h [ k ]* r [ k ] = h=W k [ K ] X [ k. for discrete systems it f .1. y(t) = h(t)* s ( f )  7 rx.
5. iii the corriplex splane..3 (01thc ztraniform from Chapter 13.12) T h e signal ~ ( t= ~ (+t 1)ej is a rightsided signal. The following investigations can be further gnipraliscd. fox J z l 3 U .z(t . By ) shifting the zero of the time axis.sponses of contzriuofirs or d. that is causal: u ( t ) . we have t o position the zero point of the time axis so that t1.ca systems.e.. rigtitsided signal.W~I a5 causal ssyncds: nition 28: Causal Signals ('15. . of the ROC' of the Laplace tiansform from Chapter 4.scr. For this reanon. It converges to the riglit of a vertical lint.for k < 0 .2).10) and Causal sz. i. we can obtzzin an ecpivalerit signal u ( t ) ..t. The transfer function of a caiisal LTIsystcin can lie ideritified significantly more precisely than orilv h i the locdion of thc ROC'. Causality aritl the Hilbcrt Transform This statcmeril is equally txue for both continuoiis and discret. Fioni tkw propertit. We have already s e m that the irnpiilse responscx of a causal LTIsystem must be a. we can derive a genrral property for the ROC of the transfer function of a causal LTIsystem.sided sigiials. hi R e { s } > U . The transfer function of a discrete causal I:il~y~tem the ztransform of a is rightsided wries.ete systems.w rightsided signal begins no carlier than at zczo Because of the close connection with cansal systems.370 15.e. It converges outsicle a cirrle in the complex zplane.[ I ) . Given separately for coiitinuous and discrete systems: The traiisfer hnction of a continiious causal LTIsystem is the Laplace trailsfor of a rightsided function of time. We will lcarn about this property nsiiig the spc i a of taiisal signals In the next sectioxis. irrespcxt ive of whether the signal is an impulse respoiise of a n LTIsystem.t + ' .1) = E ( I ) c . but is not carrsal. for the zinpulue re.gnds are szgnals lhut fulfil1 the conditzons sct out (1s. rightsidvd signals tlmt begiii no earlieli than at t = 0 or k = 0 are also IUK. i. it we extend them to cover all right.. so urrd z [ k ]= 0 . (15.
 . multiplying it with a step funtion ~ ( t ) not change tlw signal does h. We will be investigating the characteristic properties of the spwtrurrr of causal sigiials in the next section The results will be general for causal signals and their spectra.IH ( j w ) * .15) and from there it follows irnrrredia.16) This result means first of all that the spectrum of a causal signal is not changed by convolution with 1and division by 7i.15. . As every caiisal signal can also be an impulse response for an LTTsystem. l 15.jw) 2 1 1 + . wc 3”’ split (15.) (I 5.jW) ?r 1 * . } 3w determines the imaginary part o l Lhe cmvolution product and vice wrsa.t] n (15.17) 1 1 Im(H(.13) Taking the Fourier transform.1 Cont inuousTime Signals We will start with the construction of the spectral properties of continuoiistime signals. If a sigiial h ( t )is caiisal and does riot contain any Dirat impulse at. Cansal Signals 371 15.re also valid for the frequency response of c a ~ s a LTrsystems.2. 2T (15. & ( t.(t) = k ( t ). lrowever.teiy that 1 H ( j u )= n(. W (15.16) into real and imaginary paits: Re{N(jw)} = Iin{H(jw)] 7F 1 1 *W (15. together with the spectrum of the step function (9. we obtain 1 H(jw)= H(jw) 23 * [ r S ( w j i. 3W ( I 5.?5).141 We procreil using the selectivity property of tlic Delta impulse 6(w) H ( j w ) = If(. eetra of Callsal Signals The causality oi a sigiial is a property that rniibt also be expressed in the spectrum of a signal. In order to illustrate this bettckr. f = 0.2. these spectral properties .92). We can rccognise that the real arid imaginary parts can be transformed into each other by .jw)} = Re{H(jw)} a * .2.18) As 1 is purely imaginary. a n d using the multiplication property (9. it is the real part of the spectrum R c { H ( J L J ) that.
although here it is defined in the freciuencydt. aiid is defined by The Eilbert iri\nsfomni consists o f it convtzltition with arid a division by 7rq and is also itii LTIsystem.33) . ratlirr it a Etrnction of variables (here frequency) to another fiinctiori of t he same variables. tlii true for the frequcrrcv response of causal LTIsystems.20) (15. cannot given independently from each ot her .U .18) by h ( t ).(15. his yields . for exaniplc. Clearl>ithe 1ca1ant1 imaginary parts of the f1eqiieiic.19) demands some care. As the irnpulse response of a causal system is a carrsal function.20> 15.y ~espori~le a n LTIsystorii. which ic: 1 lien the I$zlbe~& tlansforui. b r k The result.We replaw h ( i ) o . Laplace aiid Foiirier trimsfonnsthe Milber t. tmnsform is not a transfc)rrnahon~ri between the timedoniaiii and frequencydoniairi.3 72 15. of and likewise the magnitude and phase.iriain. Calndatiirg the Hilbert transform by twilitatirig the integral in (15.21).21) We c m now reach the ronclusion that tlir spectruni of a causal signal has the characteristic that its real arid imaginary pax ts can be litmd fiorii eat% other by the Wilhert transform (15. 201. In contrast to thr. The correct procedure is shown in detail in 119. With the Erilbcrt transform we can concisely fomul the relationship between tltr real and imaginary p ~ t of the spcctruin of causal signals: s 5 (15.e k f ( ~ = J ) in (15. This ctperation is called Ihe FIilDrr2 truiisfomi. derived in lfiis section can be easily extended to sigrds I & ( [ ) that contain a delta impulse hoS(t) at the origin. as the dt~iominator becomes zero for 17 =. In place of (15.20) arid (E. which we had initially exclud~d. Caiisality a d the Hilbert ‘lkaiisf(~rrri an operation which is ideiitical apart from its sign.ho6(t)oeW(p) .21).h o .
convolution @ of N(e3") with Ihe coristant h [O]. This is r ~ . we h a w t o deal with the value JL[O] at k = 0.pIQ . (15.1 . Causal Signals 15.17) slid (15.2.33) and itsing the selectivity property (15. that does nott cliarigr wlreri rntiltiplied with the the unit step function: hjk] = qk. We d a r t with 8 causal discrcte sign1 h [ k ] .28) Carrying out cyclic.27).49) ailcl thc spectrum of thr unit step function (12.27) In order to firid a similar relationship betwren (75. (15. so t bat instead of (15. ] . Hilbert transform for periodic spectra.1 ~ 1.2 373 Discrct e Signals The case for discrete causal signals is very similar to tliat which vyr have just Pxamined for continuous signals.26) Finally il follows 1 H ( p ) zzz f3(&2)Q 7r .15. .2. E l k .29) as well. we can iise & yields the valtic (15.24) The spectrum of this signal is obtaiiiecl with the inultiplicaticm property (12. Rewriting as r e d a i d imagirrary parts gives Here again are Lhe rclationships b&mm real and imaginary parts that we eau simplify with a suitable definitioii of tht. Using elementary trigonometric eqiiatioris we write (15.2.38).
(15.20).22. 'H{Im{h(t)}} Tm{h(t)} = E{Re{h(t)}} Re{h(t)} = .* H ( p ) is not a rpaI function.374 15. We start with a rightsided spectrum that contains no nvgative frequency cornpone& s H(p)=O for u<O (15.%.23). the timedomain and frequeiicydomain are simply interchanged.ended to signals with nonzero mean having a Dirac impulse 2nH"J(wj . First of all we establish that H ( p ) does not exhibit conjugated symmetry (compare (9.37) (15. {Re{N(rJC2)}}. can also be ext. The properties of causal signals and their spectra can also be extended to rightsided spectra and their corresponding functions of time.ween tohered and the irnaginary part of the spectrum of a cansal discrete signal Re{H(eJ")} = Re{h[O]) f 7I. in order to obtain a dual result like (15. E ( W ) we can proceed in t1t~ same way as foi equations (15. (15 21). tiineinvariant. system.?='{H(Jw)]. and it is also the case that with discrete causal LTIsystcms. (15.38) This result. Causality and the Hilbert Warisform The Hillwrt transforrri %* is also a linem.33) (15. Then it follows that h ( t ) o .36) H ( J W )= H ( p ) .{lm{H(e3")}} Im{H(cTJ")} = Ini{h[O]} .49)). 15.21). Starting with (15.35) and are interested in the special properties at the tinie signal h(t) = . but in fact consists of both real arid iiriagiuary parts. Because of the duality. using the duality between the time arid frrcluencyttJmains.ONO their spectrum in .13) to (15.2.34) These reld ioriships are very similar to their continuoustime counterparts (15. that is limited by the imilateral spectrimi. We thus obtain results that ale closely related to those found iii Seclion 15. 'riius me obtain 7 the relation bet. the real arid imaginary parts (likewise magnitude and phase of the frequency response) cannot be given independently from each other.
From (15. The relations (1 5. from a real signal r(t) a new signal x1( t ) can he derived by (15.j . lt therefore has a onesided spectrum X ~ ( j w )The mean value of the imaginary part .39) follows ~ { x ( t ) }F 1. a certain real part Re { h ( t ) ) . This procedure is already well known.(t) t . The fiinction of time a($)corresponding to this spectrum is the Hilbert.45) Xl(f) = . transform of ~ ( t ) . Therefore. The alternative calculation in the frequencydomain is the multiplicatioii of the input signal's spectrum X ( j w ) with the transfer function o fth e system 31 HE(jw) = jsign(w) .41). Using the convolution property (9. The fuiict.19) is true: I 1 ?L{x(t)) = 1 ( t ) x ?L{x(t)) n n *t t 1 5 n ) ( . of a system 7l on the input signal z ( t ) (Fig 15.41) describes the convolution of the input signal z(t) with the impiilse response. if the spectrum has no contribution for U? < 0. tz 1 3o Jg+ 1 / 03 a!.39) and( 15. (15. it can br seen that.19) just called t and not w .j X ( j w ) sign ( w ) } .3).he Hilbert trmsforin as in (15.ion l/(nt) is then the impulse response of the system 31 and (15.19) and likewise on discreLe causal systems and uriilateral periodic spectra.ransforin of a time signal as the effect.41) can be obtained by switching to the frequencydornain. if we think of the EIilbert t. has been arbitrarily set to Ini{Ho} = 0.41) I Here the indepericient vtriablc is in contrast to (15.131{x(O} . the same definition of t.40) hold for its real and imaginary components. A shorter and significaiitly clearer way of carryirig out the Hilbcrt transform by calculating thc convolution integral (15.15. = ( (15. . it coiild have been any other value.40). Furthermore. Signals with a Onesided Spectrum 375 For time signals. also defines the corresponding imaginary part apart from an additive constant.42) The transform pair (9.70) on (15.43) The Hilbert transform S ( t ) of a time signal ~ ( t )obtained if a new spectrum is is formed from the spectrum X ( j w ) by inverting the sign for w < 0 and multiplying by .. n. exploiting duality this method can also be used on (15.41) yields (15.3.44) This method using the freqiencgdomain is significantly simpler and safer to use t hari evaluating the convolution integral (15. (15.
antl hccause of this symmetry. The signal .W I . This is where the Hilbert transform comes in. p )has i+ bandpass characterist ic with limit frequeiicies w I. z { t ) as is real. It is advantageous to tramniit uiily one of the two sidebancts.2 Figwe 15.nd contains nu further information.2).2 we t a n that a conqdex bandpass signal with a unilateral s p e c t r i m (shown in Figurell.376 15.46) w. Tlie original signal ~ ( tc )m . w a l signal .rurri is shown.2 and the desired rightsided spwtruin XI ( p ) . X ( j u ) is complex.'The analytical signal allows a simple description cif impart ant.I 1 ( t ) (15 45) has thc rcal and imaginary p't't)s Re{si(t)} = z ( t ) . The following example shows il typical applicaticm. far the sake of simplicity. Init 111 Figure 15. rcgardless of the spccific values of w l antl wa.s spectrimi X ( .It.3. can bc critically sarripled with It . filter banks antl others.ind "'2 In general. Causality and the Hilbert Tiansform s(t) ! H __ 2(t) 1 0 Tt 8 = e(t) 0 0 8 I I i Figure 15.14) requires only half tlie sanipling frequency that a rcal bandpass signal nevds. tlie lcft sideband is a reflection of lhe riglit it. ljroni Table 11.4 shows art arrangement ftx transmission of a real signal ~ ( 1 ) . Xrn{a~(t))= ' k { x ( t ) } ( 15. 'Yhe resultant signal . T now t remains to be sbowri how a complex bandpass signal with a imilatcral spectrum is for med from a real bandpass signal. communications arid sigml processing systcrns lilw inodrilathn.(t).4 only a real vainlucd spcct. so that only half the baritlwitlth is needed.3: IIiIberttraiwforrrl in the thedomain and freqrrertcydomain The complex signal of time ~1 ( t ) is called the nizalytrcal sigrml corresponding to thc. C1titptt. X ( j w ) has conjugatc symmetry (9~49). The rnissirig sidrband can be reconstruc ivrr a5 the syxnrnctry is known. It is also useful to reducp the barid signal is going to be sampled (set. Example 15.x.xz(l) has a periodic spectrnm that does not overlap. = w. sanipling of bandpass signals.r 11.
and their corrihiriatiori yields tlrr signal sl(l).Yz(f)} using a c~mtplt.5.. ~ ( t > . which has a onesided spectrum. Signals with a Onesided Spectrum 377 .1.3.:bandpass RP. w Figine 15... x ~ ( tand z l ( t )m i s t be combined accordingly and ) .3(t) and x d ( t ) produced are both complex signals. The real part Rc{zl(f)} i s the real original sigrid . The signals s.40) 2F 1 and the sampling interval 7 = ‘ w2 . with trarisfrr function ( 1 1.(t)} and Ixri{.w1 .4: Signal transrnissiorr with thc Hilbert ttransformazfsor be obtairied by filtering hoth of the real signals Re{x:.
For each part.3. each sampling value has a real arid i. txansform of the even part x e ( t )and the odd part z o ( t )have? . Some properties of the Hilbert 1 ttransform y ( t ) = X ( x ( t ) }= . it is clear that in one branch only the real part of the impulse response must be considered. Exercise 15. give the phase shift between the input arid output signal caused by the Hilbert transform Exercise 15. Howcver.3. we inay use analogtodigital converters with half the sampling rate aiid we need not pay attmtion to the relation between the band edges w1 and w2 for critical sanipling. The required number of real sampling values per time i s thus the same.3. imaginary or complex? What symmetry does the Hilttert.2 For w g > 0. consider that in oiir example.cgz[k] 1 b) ~ [ k ]u [ k } ~ [ k ] = c) ylk] = a [ k + 1]x[k] d) y[k] = sin(n .378 15. the sampling rate is reductd by a factor of two. sin U . As the input signal of both complex handpass filters is ) purely real. a) Give H ( j w ) e0 h(t) arid sketch the magnitude arid phme of H ( j w ) . s [ k ] ) ej y [ k ]= 4 2 k ] Exercise 15.1 W'hich of the following discrete systems are causal? a) y[k]= ClZ[k i 1 i. Causality and the IIilbert Tri2rlsform for ~ ( tto be recovered. and in the other only the imaginary part. The irripiilse response ?t of the Hilbert.jw). txansformator is indicated by h(t).3 Consider the real signal z ( t ) with spectrum X(. Howevei. $ . c) cos q t and d) b) cos 2wot in the frecluencydornairi. Compared to critical sampling of real valued bandpass sigrials (see Chapter 11.4).*:c(t) are to be investigatcd. How does the Hilbert transform affect / X (jw)and arg{X(jw]}? 1 b) Is g(t) real.4 all sampling values were real. calculate the Hilbcrt transform of a)eJwof. while in Chapter 11. Note: see Figure15.tii imaginary part.
Which spccial property exhibits X ( l ( ~ w ) .7 5.e. ( Note: Consider which transform corresponds t o the required convolution. Which property of the = Kilbert transforni of real signals can be derived froni this result? Note: see Chapter 9. zl(l.o ~ ~ ( t ) ? o7= d) Calculate X . Exercise 15. ( j w ) forrnally irt depcndciice on X ( J W ) . c) Determine the rela. yZy(Q) f".4.3 c) Let r n ( f ) x ( t ) + . and its IIilttert transform i ( t )= 3I(e(t)}. sin(ln/&)].4 Given the signal ~ ( t ) 2 = w 1. cos(w0vt) + b . d) Compute the Hilbert transform of the Fourier series [a.9.et X l ( j w ) be the sprc:trurn of the real signal sl(l) and X2(jw) = (1 sign(w))XI (jw) onesided spectrum of a signal :EZ ( t ) . Note: refer to thc% Parscval theorem in Chapter 9. b) Determine d ? ( .8 aiid 9. b) Determine the relatioii of the signal ciiergies of energies directly in the frequency domain.tion of the signal energies of x ~ ( t und zz(t)by a consid) eration in the time domain.erminc z2 t )in deperideace on % 1 ( t ) . Exercise 15. Exercise 15.7 A sketch of the real function P ( J w ) sllomm.15. i.5.e both . Note* consider o Figure 15. ~ ~ ). Note: Vse the multiplication tlieorem of the Fourier tramforin.8. s ( t ) y ( t )d t . a) Calculate arid sketch X ( p ) oor(t).t uJ9t n . Exercises 379 c) Calculate the crosscorrelation fuiictioii of the (deterministic) signals z t ) ( and y ( t ) for z = 0. with WO > 0.5 What is tlie connection between the energy of a signal ~ ( t ) the Hilbcrt transand formed signal Z ( t ) = 7 f { x ( i ) ] ' ? Assume that the signal has no DC component. ~ i ( t ) .. Exercise 15.cosw. Chapter 9. the + a) Det. which is part of the transfer function is . llse the result of Exercise 15.6 1.)urid ~ 2 ( t ) Stnt.o ?(t) graphically and calculate r i ( f ) .
4 Q(2.l) and Q ( j w ) show'! q An audio signal with thc spectrum S ( ~ Jis to be trai1sniittc. x a) Thr signal s ( t ) is nrioclulsted by cos(wot).38 0 15.ornparect to S(SUJ)? c j To rc'inovc the disndvantagr foimd in part b).( t ) 0 . the signal will be trarisrruttced according t. the signal spectrurn i s 0 for \ U / > ~1~ = 27~. 4 kHz. in the spectrim of it red signal. . 1 €(O) = .8 r"(:p~). 2 b} Firid the imaginary function Q ( p ) (15.jw) oi a causal system.18). I>raw the speetrurn of SEn&w). known as single sideband modulation (SSB) takes advantage of the rcduudaney prcwni. This method. Causality and the Hilbert Transform H(&J) = P(.o tlic following representation. a) IMerrniiie the corresponding iiripiilse response Ir (t j 0c H ( g w ) . . with SOW: c) What symmetry do p ( t ) 0e Exercise 15. In ordex to bave ) bandwidth.jw) + jQ(. b) How much baiidwidth does Snf(jw') use c.d.m cos o'ot Giw the spectrum of s h f ( t ) .
4. n r ( j w ) compared to S’(jw)? use 381 .15. Exercises ct) I3ow much bandwidth does S ~ .
This Page Intentionally Left Blank .
optical. 1 There me various possible ways of definirig the stability of a system. Then we will restrict ourselves to caim11systems and coiisidei stability tests using the pol~xerodiagram. Among them. For a discrete system the signal ainplilude must stay within certain limit5 as defiiied by the perrnittcd numerical range of. another important criterion that must be met for a system t o be implemented is stabilitj. BlROstability is particularly useful for LTIsystems. 2 zs s. hydraulic or pneumatic systcms. The first thing we will do in this chapter is investigate the connections between stability.the coinpiiter being used. First of all we will introduce the concepts boimded functzori arid bounded series.A5 wc.uficien. ' d k . Fiiially we will disciiss sonic typical uses of ms with feedback. the conditzon less than U fixed IimPt (16.11 as causality. if continuous systerris are to be rcdised based on the fundamental pliysical principle of conservation of energy. and the input signal i s bounded. rnechanical.1) lz[k]l < M < w. ~ ~ o ~ ~ e s ~ o ~ d a n ~ l ~ . We will again be dealing with continuous arid discrete systems a the same time.2) n senes to be bounded. then the output signal cannot grow beyond limit. 21: Bounded function. the frequency respoiisc and the impulse response foi gmcral LTIsystems.Slt. II a system is stable. This condition niust be fulfilled. for example. With these definitions we can define stability: . electrical.t for (16. bounded series 2s A functzon as s a d to De bounded when 22s muynstude for nll lime t: [z(t)j< Ml < 00.
3) We will iirst sho.wi upper bouiid h r ly(t)\: Note that LIE inipulse rcsponse musi be abolutely intcgratable.1 ~ O ~ t ~ ~ ~ O U S bta From the general definition of BIBOstability. In order to show this. From these. Stability and Feedback S y s t e m This can h r nmtly expressed as: Bourided Input + Rounded Output. This stability definition i therefore rallrtl RIBOst ability.tbility.5) . '1'0 do this we form thp ntagnitud~ the oiitput signal Iy(t) 1 using the coiivolution integral. the following colidition for the bility of contkinous LTIsystems c%tii derived: be A contirnious LTIsystem is stablc if and only if its impulse responbe can be absolutely integrated: (16. stability conditions can be given for the irnyulse response o LTIsystmns. of Takiiig a bounded input signal z i t ) (16. m7c will derive this separately for coritiriuoiis and discrete syst~ims. we tonddrr the rat her unpleasant botmded inpiit signal (16. wr can directly iiiid . we can also firid conncctions with the frcqueiiry f rrsponsc a i d the transfer fiinction. . responsse h ( t )that call be absolutely intcgratett.vlrthat this condition is sufficimt for HIUOst. s As with caiisality.1.1) and an irnpulsc.3x4 16.
2.5)). the fi cqucncy rtsporw o f R stnblc svstein m i s t riol h a w m y disc oritirriiities or ringularitics. arid is thv saim as thc L a p l a ~ e transform on the it c m be analytically corrti imaginary axis s = jw. (16. d.2.7). That mcaris that for slable systems. E IR . if the impulsc response can be abwluLely iritegratetl. Example 16. stable system cannot h a w ariy siiigiilwrities or discontinuitirs ~ iscrete Systems For discrete LTIsystems: A discrete LTIsystcui is stable if and orily if its inipulsc response is (16.7) I ~ T h e prod is exactly as for continuou5 systems. the frequcricy response of the stable systerri rmist exi Because the Foiirier integral is 1)oiindrd (SCT (9. JVe will clarifv the use of the stahilitv criterin (1 S.. TL can be slion7ii in the same m7ay liom thci cxistenre ot thc frequel~c.16.1 Fhr the simplest clxaniplc we consider a system . with irqiilse rcspoiise h ( t )= ~ . Impulse Resprise arid Frequmcy Response Curve itrid calculate the value 385 The result is only finite when thc inipulse response can be absolutely intcgrated. BIBO. '12ic wnse of' a. As we discovered in Chapter 9. tlrp imagiiiarv axis of the splane is part of the rcgiori of c o i i v q p i c e of the system fuwtion. using (16. Correspond} ingly.8) .y rcspot the Foiiricr translorrri I I ( d 2 )= F*{h[kj} tlie iinpiiBe response h [ k ]agrees m i t h of the transfer function H ( z ) = Z { h [ k ] on the iiiiit circlc of the :planr. the Fouricr transform H ( j w ) . ( f ) rx.." ~ . othcrwisc l g ( t ) I WOLIMriot Le I)onnckd.1.3) tor contiriuous systems with a k>wexsinplei Stdbility is showri in tlic s a n e way for discrete 5ybltims.F{h(t)}.
and with (16. In ordrr t o confirm khis. the impulse response takes the form of the step function h ( f ) ~ ( f ) . finite vdue or grows hq7ond limit. the unit of tiinc in (11 3 5 ) is chosen so that T = 7r for simplification. The baselines of the trianglcs each have width 7r arid heights decreasing by 1/t ‘l’lic infinite sum of these triangle arm i s a lower bound for the area uritler the iriagnitude component ofthe itnpdse response ( h ( l ) l : . The ideal lowpass is thrrrfore not c.10) W e iiow have to deterinine whether the integral has a. wc investigate whether the impulse response can be absohttely integrated: Ie‘’’E(L)l tit = bW encdt = c1 1 for a > 0 (16.11.2. We now want to invesligate whether such a systeni caii be realised.2 When we clealt with the sampling theorem in Chapter 11.d a system with a rectangular frequency respoiise and a sint function impulse response (11. (16.4. Stability arid Feedback Systems To deteririirie the stability from the criteria in(16.9) we can say the following about the sgsteni stability: an integrator is not BIBOstable. For m = 0. it can be immediately seen that the impulse response of the ideal lowptws is a bilateral signal. we 1isc. This is also called an d e a l lowpass filter.y estiinatiiig the area under l h ( f ) l with a series of triangles whose areas are all less than the area of the individual sitlelobes in Figure 16.3). Considering the impulse response in Figure 11. incrcasing time ovcr all limits arid thus violatcs thc condition for BIBOstabilit Example 16.1.386 16.1.4. As the frequency iesponse of a stable system cannot ha~7e any tlisront iiiiiitie4.3. = We recognise the impulse respoiise of ail integrator froiri Chapter 8. we suspect that the ideal lowpass is also unstable. This is not surprising when we recdl the response of an integrator t o a step function ~ ( t )The integral over this bounded input signal grows with . We do this b. and the impulsc response is h ( t ) = si(t) .35) as an interpolation filter for sampling.iusal.9) c otherwise m arid obtain the result: the system is stable for a > 0.
and reconstructing contiriirour sigiials fruiii samplPs. The ideal lotvpass rmriot be prrciscly rediwd lmt can bc apprtmimateil if miiie coritaehsions can be accepted. . for example. BIBO.Noncritical sampliiig. limiting spectra. The ided hWJ2aSS is therefore neithel c i ~ u s d nor stable. A n cxarnple for this coiild be: 0 .16. filtcr can be tipproximated Ttic impulse resporise of' the ideal lowwell by i i causal system if it is shifted far eiiough to t h c righl. which tihows that the impulse response of the ideal lowpass cannot bc absolutely integrated. arid does riot oo~iverge. The answer is typical from tlic point of vicw of' systems theory: o The ideal lowpass filter with an entiiely real rectaiigular frqiency plot i s a very simple cority)t. when they arc innpossible to impleiiirnt. Q Af'ter this rxample we have to ask whether it is worth examining such idralised systcms.'rhe irripiilsc rqmrrse eii drcays more quickly and therefore can hc absolutely integrated and stable behaviour has beeii achieved  Permitting a timedelay.1.. Oversampling allows the steepn of thc edges to be rcduced (see Figure 11.13). It allows many operations to be coiisideied in tire simplest form. Iiiipulse Response arid El'fequency Kc>sponst:Curve 387 The slim itself is a known harmonic series.
Necm the irnit circie iiitxst be par. 'rhereforet . the radius of the bolder must be less t l m I. this mcaiis tlmt the bortlei rrriist lie to the lclt 01 the imaginary axis.I ~~~~~~~~ ~~~~~~~~~S 16. It ~ ~ ~ i ~that eit must br pos)li s sible to ~ 1 ~ s ~ ) f ~ ~ t e f tlic iiilpulse response.5. circlc arouud the origin of the zplane.of lhe ROG fur a st&k syskrrn. . ~ ~ with ~ ~ a real t Now we look ag&iiiat tfip property of stability.1. 'We will look back a t ahoiit$ i o ~ the ROC of the Laplace transform in Chapter 4. howcvrr.tginary axis.. this iritegrde ~ nicans the irvragiiiary axis of tlw splane iiiust be part of the regioii of corivcrgence of the. As LEiib must lir to the riglit of a border paidlel t.me time can be recognised by the location of their siiigiilarities in the corriplex splane. the traiisfw fmction of a causal Ll'lsystem converges in a regioii of tlie splane that lies to tlre right of the s ~ ~ ~ ~ the gmctest r ~ part.5). The RO@ of' the Laplace transform for rightsided signals lies to the riglit of' a line in the s . in the left. in t k left half of the a r ~ i e lie splane. Whcre then. 5% will first give thcir properties in a gmcrel f c m n and then sprcidisr the iesnlt for Iil'Isysteins with riitwna 1 traiisfer Smictitms. and since the sirigularities caniiol lie iu the ROC. 16. do t'tie s ~ ~ i ~ of the~trarisfer functiort. Therefore.2 Discrete Systems For discrete systems we look at the properties of the ROC of 12ir 2lraiisform.o the imagilimy axis. F~JT a causal discrete LTIsystem.I Continuous Systems The 1Cc)C: for thr trarisfczr function of a ctmtiiiuour. the ROC of the transfer ~ r i ~ Ht( z ) lies n ~ ~ cmtsitle a.. Chrawlity mems that the i m p l s e rcsponse is a sightsirled sigzial.388  16. \/Vc havet alreadj said !hat. thcy m i s t I)c inside this circle.p h e . to find simiiar results. itre causal and s t a l k at the s. traiisfer function. Wr can summarise this insight with t hc following statcmerrt . Stability ttrid Feedhack Systcrxis liTlisystenis (hat. half of the splane (see Figure 3. wliich is parallel lo the im.(.~ ~ Since lhr ROC is i ~ ~ r ~ t i ~ lie? free frorri s ~ n ~ ~ ~ ~ theyi must s ~ t o the left of the Linc.2.3. LTIsysteni gives 11s important ~ ~ ~ f o r ~ ~ a tthe camaiity artcl stability of tlic system. .2J.
Causal Stable LWSystenis system function W ( z ) for a causal.J W ? . ~ I response to just mhat we expect for a stable system. charitcteristic frequencies mtl thus hi$% influence on the stability of t h system. tlie charact eristic fr ecluencies of the system.he rwi( circle of the zpli3ne.d ~ the ~ a i ~ ~ .2. < 0 is in the lef%h d f of thc plane.3 which shows that only poles in the left half of the splane coxrespond to decaying cqmrcxntial illalions.  _. discrete IXLsystem must lie within t. ( sigiial from Chapter 7 .ernal tcrrri can be represented (set. Their sixigularities are siinplc or inultiple poles ilial cfelint.14) The condiliorr for stability. It describes the part of the output sigiial tltat is caused by the initial slate of the system. The Laplace transform of the int. We caii describe t l m e systems with polezcro diagrams. ing characteri~~.3) a1 = (T. in the t i i ~ ~ . means that. 1x1 functioii H (s) for 8 caiisal and stable roritinuous left half of t lit. The location of the zcros in the cornplcx plane does not influence t Irt. stable. 31 83 f The genctal statemeiits CRU IN>inndc more precise for svsteizis for which the transfer function i s rational. (7.ic frequency decays tec the liln c'h' = lirrl p a ~ . splane. tlic real part CT.92)) by pwti tioris with poles s. This IS ctcar from Figure 3. no ~ . the iiiitial coIiditions decay5 with time. that all poles lie in the left half of the spimc. c)"'' f tca z0 I (16. In the timetlornein this coiresponds t i t the sum of the 1's characteris< freipencics..1. 3 3 .1 'PO illustrate this property w e m i s t recall the interilitl term ylzlT6 ) of (be output. The decay of ttie inttcrrral tcruri mc that ttie iiiital state of a system dors riot define the system Feknviour.16. /. whitli we ic have only written uiit here for N simple poles: (16. a s long 22s one is prepaxecl to w a i t long enough.
We will only br ronsidering one test for continuous and discrete systems becai~se tests all have essentially the banie effect. We can bee inixnediately fiom tlie locatioix of the poles tlmt only the first and third conlirruous system (likewise discrete systeixi) are stable.ional fonn is determined from a differential or cliffercwe ecpat.5). they jnst determine whether all poles lie in the left half of the splane (or unit circle ol the zplane). The stability conditions can alsn be illustrated to aid understauding. only the numerator and denominator coefficients arc: obtained at. the zeros of the denominator polynomial must be determined. The internal term of the system ~esponse here only decays if thr corresponding poles lie within the unit circle.22 .ybtern lie within the unit circle of the xplane.the first three continuous systems and tlie first. the . If. To avoid having to perform the nurnerical senrch for the zeros by linnd. but for higherorder systems iterative procedures arc necessary. and the system will always be ematabl~. po1t. first. 16 2 shows four polezero tliagrnnis of both contix~uous (left) a i d discrete (right) system. as in Figure 13. Again. In order to check the stability with the location of the poles. The ROC must be chosen so t h a t it lnrludes the imaginary axis of the splane.390 16. This can be given in closed form for polynomials up to the third degree. t h i s t. a series of easy the t o perform stability tests were developed.6.ion. Stability and Feeclback Svsteriis 16.3 Figarc. IIere we can also see the connection betweerl the locsation of tlie poles in Che complex frequency plane arid the system's characteristic frequencies in the timedomain (sec Exercise 16. 2 Discrete Systems A rorrcspondirrg result tiolds for discrete systems: The poles of the system fiinction H ( z ) of a causal arid stable discrete LTIs.s lie directly on thc iniaginary axis or unit circle. Example 16. hom7ever. Instead of cal(*rilating iridividiinl pole locatioris. biliiteral impulsc responses are permitted . <)I' the unit circle of the aplane. but modern computers can do this easily. the initial states have rio influence if one waits long enough. second and fourth disciete system are stable. Stability Criteria If the trijnsfer fiinction of an LTIsystem in rat. Once the digital computers needed to carry out these procedurcs m w c unmailablc.ecornes irnlmssible.systems which are not causal .
arid directly tises t h e cocKicients of the <lc>nominat.3* S.16) .2. It does not irivolvr searching for zeros.2: Polezero plots for stalile and unstable systems 16.15) so that the numerator polyriomial (16.or ~ ? o l ~ i ~ o ~fiii the trassfcr function.16.2. Causal Stable LTISystems 39 1 t. i ial To use it w put the transfer function in the forrii e (16. colltilluous Systems For continuous systems we will describe Hurwitz’s stability criteria. t I I A Figure 16.
a Ilurwitz polynomial if all Hurwih deterniinarits arc positivc. fiilfilled if all orders sn in &(s) exist. Q.N _ I (16. Tlic bzlzneur transform is a sullabl~ way of doing tlris: (16.3. lilit~iliustratc how it is carried out in Example 16. 2 . and Nurwil z’s stability criteria dctcrniiiie wliel. To hrrnulatc the necessary iitrrl sufficient conclition. > 0.4. The test consists OP two parts: A nwessary coiidit. ~ 16. * .. n = 1. . 191. > 0 for p = 1. as otherwise tlnert. It can be only ht. For A”= I arid N = 2 this cwndition is sufficient and tlie t for N > 2 the following conditions must also be tcsted.. A rdatecl procedure Rotitli’s stability t r h t . are positive. dihcrete syste~ns...20) That concliides the Iliirwitz test. whtre we can carrv out a stability test for continuous systcms. . Stability and Feedback Systems A polynomial i s a Hwirirt~polpiianiral if all of i t s zeros have a negative real part.302 16.18) (16. the Hvru12tz deternununts for / I = 1 .17) v. N are set up: ( 16. A{.2 Discrete Systems Foi.her this is the case..21) . ‘The system is htablt. if its numerator polyrioinid Q ( s ) is a Hurwitx polynomial. must be tkterniiiitd whetlic~ poles lie>within t he unit it all circle We will fhcrefore be using a procedure which maps the cornplex Z .2.19) i.2.ioiifor a Hurwitz polynoniial i s that all eoe%cicnts a . . would be a coeficierit u n = 0. .~ ~ R I I P onto the corriplcx splanc.can bc found in 123..(16.
y critcrion for a discrete system with trausfcr function I I ( x ) is now: . < 1 follows t~ < 0. On closer ~ ~ i ~ ~ ~ ~ ~ i procedure is still Raw~d: point x = 1 is Iiowt. biliiirar transform (16. t o test t h i s point Q(1) well.21) in comparison with &hcr traiisfoorrnacions that i m p the area wi&n thc unit rircle onto tlie left. The stability of the tliscrek systeiri (16. the zeros rnrrst lie in the left halt of Ihe splane and tlw zeros of Q ( z ) nirist lie in the unit chck of' the 2plane.Zfj). the o tltc point s = cx) (bee (16.vw.2.23) is guaranleed if the zc'r'os of the tfmorninator poiynoniial Q ( x ) lie in the imit circle of the zplane. Causal Stable L"1Systeiiis 393 f 16. this iiiirig tht.t her the mimerator of G(s) i s a Humit&polynomid.so we test w11t. St. lf it is. for () examplc.16. The biliiiear transfor~nation (16. half of the splanc has the advantage that @s is a rational h i c t i t m and we can use the Hurwitz lest.24) Zeros of Q ( s ) (and so also of ~ ( z )can only arise from the nunrerator polynoniial ) of @(R). To d i ~ & ~ j v f ' f o r m the the ratioid fiinction G ( s ) from Q ( z ) .22) For 1. our o I ~ . so it is not iricluded in the iftirwitz t.
+ T h r a system is unstable.1) ~ ~ . zeros =$ Q ( z >has zeroh will1 jzj 2 1. (s .' ronditioii for i~ ~ ~ ~ l .3!)4 16. s ~ Wurwitz .25) and froin here obtain the syskm fiinction Using the bilinear t r ~ i s f o r ~ i L t ~the .q4 .de~1omiK~alv~ ) becomes the partial fracon &(2 tion Qis): O(s)= Q(%) 1.27) W e mush T ~ O W test the polyrioniial h parenhhis in the ~ u ~ ~ of~g (d ) ~ o r s to Ilstcxtain whether or not it is a lIuiwitz p ~ l y t i ~The necesiq. In order to deter~~~~er mine its stability we f i i ~ s t read from the block diagram: (16.114 (16. Stability and Feedback Systems Figure 16.3.3 s11ou. t h zeros of the deriominator polynomial &(z) can be found using a ~ coniputer.17) is not fulfilled as nut all o f t Itt twfficients arc positive.86678 . am1 t h i s confirm the result we obtained using the bilitiear ~ ~ a nand the ~ r ~ tcst..rwxsive ciiscrete syst. Elurwitz pdynomial (16.466'7~~3 .1. We can now make the following dednctions: + Q ( s ) ~ i a s with Re{s> 2 0. ==+H ( z ) has poles with IZI 2 1.s a f o u x t ~ ~ . Of tourbe. Tlw result>is There is a zero at := 2 wliich is oiitside the tiriit circle.875 c (.eni. 2 .
16. Fecdback Systems 395 Figure 16. Tliej must lie ~ in the left half o the splane. T h e feedback principle is used in illany arms. Coatinuous system that havc no zeros in the right splane are called ~~~~~~~~1 phase ~ ~ ~ s ~t o~r ~ en s ~ o ~discrete systems with no zeros outside of the p .the closed loop i s then (16.3: kourthorder recursive discretetime system feedback c m be described by a transfer function.s).3.15.4 shows the situation again.15. Figurc 16.28) The poles ofthe u7hole fcedback system H ( s ) a r the zeros of G(s). but in contrast to Figlire 6. f liowever. TIP poles of G(. 3. C ( s ) (01 G ( z ) )has to hc minimal phase and additionally may G'iz)) & riot have any zeros on the imaginaiy &xis (OK on tlic unit circle) of the complex frequency plane. The transfer function of.1 ~~v~~~~~~ a ~~~~~~~ Using ~~~~~a~~ In Exxrnple 6. i~~~~~ly~ unit circle are also called t i i i ~ i n i phase systems. but the i ~ € o r n ~ a t ~ o n the ~ ~ d i ~ r i ( f ~ ~~ c~ ~~ n~ f o r w a d a d feedbark aboiit transfer n t of the o s paths is lost (in Figure 6. have no influence on the stability of tkw feedback system. both natural and technical. so that stability can be assured. We will show thrce typical problems that can be solved using feedback.15 these are F ( s ) a d Gfs)). thc feedback path coritains another change of sign. as showii in Figure 6. For thr iiiverse sptem ~l (or ___ to be siablc. .9 we saw that feedback can iiivert the transfer function of a system.
feedbacl<(illnstrated in Figizrr 16.r with a K Figure 16. ble plant can he coupled with r a feedback system. An anip1ific.396 16. to yield a coinplete system that is stable. The rcsinlt iiig freqixency respame is nearly constant..3. thc fofor. This lr d system i s also called the cmis ronstructcd so thtxt. It is important. We ill now he coalsidering whet h ~ an nn.lr.electronic iiiiqAifier. Some of the amplification is t h w sacrificed. of coiu'w. so that the negative )/ feedback systteriihas L: siifficicntly high amplification l/li. for cxample.3 Using Feedback t o Stabilise a System In many control system applications. for audio usc.4: Structure of the system G ( s ) with ftdback loop S ~ ~ the ~ t ~ i ~ ~ Wheii c ~ n ~ t i i i ~an. The system behaviour iizodified by couplizig a second m to i f . .5: Sniootliing the fieyuency rcqmise with fredbacli very high factor of a n ~ ~ l i ~ { /F(jd)/> i1can he built eabily. that J F ( j u ) is large enough ihat the J ineqiialit~y/ K F ( ~ J >> 1 is also true for small valixes of I<. by conuectirrgthc nirtpul Itack to the input with a gaiu fwtor K .tnged by siinplcl mcans. wiially as it bnckwarcis ng a closed f ~ ~ b a cloop. the principle t ii~~ o f riegativc.cvard path (the 'plant') i s a large teclrnical iustdlation that! clrtiinot IP c. . its propcrtieu can be changed without too much effort. Stabilitv avid Fkcdback Svsteins Figure lfi.5) is very uscful. it' strong variations : ~ ~ i o> ~ in the freqiieiry rcsponse F ( j w ) can 1x2 tolcratcd.
. (shown in Figrare 8. hut t h i s time choose U < 0.16.29) Fecdbark WIF dso have thcx oppositc cff'ect: a stable plant can bc clc>stal)ilised by feedback. Hcrc w e start atgttln with Pigme lC6. The transfer functiort ctE the feedback system i s / unstable for a > 0 Figure 16. It is thcrcforc unstable.ts tlic firstordrr transfer function aid R pole at s = a > 0 in the right half of the splane.3.24) with a ~ I ~ ~ l i ~ c a t i ~ ~ I<. a Pcircuit.6. Feedback Systems 397 mTewill start with the firstorder plant from Figure 16.6. The forward path of h e systerxi h.6: Firstordvr plant with Pcontrol (16. W e choose for the feedback path.
398 16. reaches thc s right half of the plane. * Negative feedback stabilises a syhtcin. jpole~pairt on the iniaginaiy axis. The tmrisfer fmiction in the forward path b lilfs) = .30) s2+u has two poles at s = k&Z and is therefore unstable for all U . Figure 16. c%l. K < (1 The infiuenrc of the Pcirruit and the sign appearing at the ~ ~ i ~ x ~ ~ a ~ i ( ~ node in the feedback.11. This Leads to an overall. U E I R (16.v and Feedback Systems b > 0 And f i < 0. as the coefficient of the linear term in Y iu cqual to zero./&it. one in tlie right halfplane. transfer fnnctitm for the system + + . Stabi1it.9.. This can be confirmed by a Ilurwita. Tlie pole of thc system without feedback now lies ix~the Iefl. or~e ~ ii ~ We cari show this with a. Clearly it is not pcnsible here ~ a ~ d troniove both poles into the Icft half of tlie splanc using Pcircuit feedbadc.31) is not a Hurwita polyriomial.path on the system as a tvliote cari be mnmarised as follows: Positive feedback dcstabilises a byst em. Figurc 16.8: Root lrrcus of Figure lli.10 shows the rook locus for a < 0. ~ ~ n ~ o r t u ~ a t c~ ~ ~ ~ h ~ r systems rthe r e l a ~ ~ i o nare~riot~so simple. the poles just move further away. Figure 16. The system c m be snccessfully stwbilisrd with proportionaldiffercrllial fwdtfaclc as in Figure 16.6 for n < (1.8 shows the corresponding root locus. For K < 0 nothing chijnges. half of tlie splane. Using a Pcircuit for fc.rdback the overall Iransfer funcriou is obtairred. For K > 0 the poies move along the ~ n ~ a ~axis~anda for y > ( . For I< < 0 it ~ n o w to the right and at K = a. The denominator polynomial s2 a K h in (16. For K = 0 all of the poles of the system lie on the real axis. k s t . secondorder plant depicted in Fignre 16.a ) / h they form a ~ i ~ Er' cornplcx c ~ ~ ~ . for l i .
10: Root locris of the control circuit from Figure 16.33) Ii’igure 16. if the constant K 1 and h zare clioscn so that F bK2 > O and n f K l b > 0 . ( 16.t.9: Secondorder plant with proportional control Figure 16.11: Secondorder plant with proptil‘tionalcilfferentialcontrol .eclback Systems 399 unstable for all a Figure 16.9 for which the denominator polynoniial is a Hurwitz polynomial.16.3. E.
2 Show that (16.c(f) is to be iriwstigated. a) with the biliizear traiisforiiiation z = 5 and a.1 Tn stability of a system with irnpiilse rcsponsc? h(t) = t”O tc It sin(5nt).. Givc both p c ible KOCs of N J each whether it is st.tetl’? b) Check tlir location of tlie poles. xercise 16. For what values of a is the b) miiic the trariskr lunction H Z ( z ) of a second d IIz( cc.61 with 0 < \ U \< 00. a ) Giw t h c transfer function N ( z ) = x(z) of the system. sigital. Stability aird Feedback Systems Exercise 16.7) is a iiecessarg and sufficient condition for the BIBOstability of a disrrtlte ILTIsysLern.) = 1. giw a mothratioii that r2 discrctc system is stablc if its poles lie within the w i t circle of tlic complex zplme (see Section 16. Zxercise 16. a ) Dcteriniiir If.tblc. ( z ) . .4 A system is giver1 by thc differcntial equation yrk] .I] t.3 ten1 is described by the following diffec.400 16.( 2 ) = and the corresponding polrwro diagram.21 = ~ [ k ] .2).v[k . b) T thc spslcm stable‘? This slionld be tlcttrrriined r a ) with tthe polezero diagram.5 Uriiig tlic internal coiiipoyicnt of the out piit. stability test for continuous loystcms.rrnitia1equation: . a) Can the impulse response be absolutely iril egra. s 3 3 system so that cl clet wmine for Exercise 18. Give 1 he ROC of I I .$y[k .2. c ) Is the system minimal p h w ? tl) 1s the systcrri causal:“ Exercise 16.
s and G'(s) = ~ 2T'O(s2 i 1) .'21)*? Exercise 1 6 3 A co ml loo] with E ( s ) .9 Consider a s y s t c m a) Is $1 viitli transfer fiuic~tion ( s ) = H 1 sz t 1' H ( s ) stable'? .. Exercises Exerciso 16. .ed to on the splane wlieii Iransforrwd hy tlic bifincar transform (16..16.is given S Exercise 16.6 Test the stabilil y o f the sgsterr~ 401 Exercise 1(3.4.7 Where is the left half of the unit. circlc on the: xpla~ie projcct.
he Hnrwitx test.Fox what values of Ii' does the ~ ~ ~ ~ ~ i l siic. Note: consider t. 1 H ( s ) is as in Excrcise 16.d. can the system then be s t a ~ i ~ i ~ If d ? Lor what values of Kf e yes. Is H ( s ) hca'nle in this b) To stabilise H(Y) proportional feedback is prwidccl by G(8) (with real amplification K ) . Can the system be stabilised'? b) If G(s) is changed to provide differeritial feedback with K real. Exercise 16. .ceett?a r ~ is~ti Exercise 16.2s 45 ' a) Draw the root lociis for G(s) = h . Stabilitv and Feedback Systems S1 will b) now Iw iiicorporated into a feotlback loop. although F ( s ) = s2 1 .402 16. X(s) Give the net*: traiisfcr functiort Hr i s ) = a) For what viiliips of #$. I is the feedback systcm srable? I d) Test the stability for K < 0 with h r ( f ) .10 a) D r w tbc palezero diagram of CiW? H ( s ) when G ( s ) = 0. 1 h w the root lociis of H ( s ) for 0 < h: < cm.10.
and torrn derivatives ancl integrals. Shis was all possiblc bcc*ause we l m l assiimcd that every s I had one definite value at avenj’ poiiit in time. for w2iich a significantly simpler calculation i s possi blt with cxpertetl valuts.inuous and discrete signals that. For discrete signals summation is iised in d of irittegration. but this would not lead to a tecliiiically realistic solution. 1% could calculate the valoes of the signal. which means that a signal has a ksiowri unambiguous value at every poiiit ill time. which will also be discussad. signal waveform can have aii uiiprcdictablc val is t h c b r efore rtid[)IT1. It ~ ~voulclbe theoretically possible to desrribe the sprcch signal from Figure 1.1 the properties of the human vocal tract by superiiiiposirig various waves. a. or ‘deterministic’ quantities. random signals must he described by nonrandom. or Sigriiils made np of c ascillations.1 wit 2. It is cornpletely impossible to assign functions to iioise signals. . W found integration e vcry uscful. This can be prrtorrn. foi convolution. add and subtract signals. by an iiifiiiite Fourier series. A signal ctui also be deterministic when it cannot.d by the socalled expected values. ~ n t wc~also u s t d l complex iiitegrrttion for the inverse Laplace and inverse ztransfornris.1 at Are Ran The signals that we have been working with 60 far are called ddernimzstzr sz. All of tlicsc forms for describing random signals will be introduced for continuous signals. however complicated that formula might be.gnala. Then we will cleal with stationary and ergo&. clday signals. he described by siinple mathematiral functions but i i i s t e d for example. a i d the chapter concludes by cxteudiiig t h t concepts to dibcrete rantloni signals. does not actually help much.1. M M Lsignals that occur in practise do not conform to this assumption. raiidoni processes.e All of the cont. we have considered so far in the timeclomain arid freqitcnrydoincri W P L signals that could be described by xiiatli~ ematical luncIions. To deal with system inputs and outputs in the way we are nsed to. An important clabs of exp tl values arc correlation functions. A i i ~ w coiiccpt nnist be found to represent such irregular pro Just unclerstwding that. md that every signal coiilil he described by a mathematical formula. which arc iritrodncc~din the next section. for Fourier arid Laplace tiansfornis. .
a frr~ii ~~~ x er distort ion or tf.x z ( t ) .1 as the noise of an ~ ~ n r ~ larid e r i ~ c*alctrlatcthe response of a postconnect ed sy In.ed irt a norrnal way. for examplc.havioiir. with an output sigiiai calculated from i t .pcctruni or outpiit signal is again it randoni v:iriablc whose behaviour we caiinot rriakr m y genmal statements about.aplare or convolut ioii irifegrals of t h t random signalh exist. or whether the iiicthods nsrd to cdciilalc~the sp funrtiovi art> weii tlefined Even whc t existetice of an integral is certain. it is inipossiblc to know the miise iignals in individual amplifirm in advance.rz ( t ) .siochastic supals or runrlorrh szqnals. as another amplifirr of the same kind would produce another noise signal.lrrmaI resistawe noise.3c14 17. In oiir exampie it2nieaiis that. When the signal form it.eiit ~ o i i ~ ~ ~ ~ t in I. is tiic same for all amplifiers of a particrtlar iiiotlel.r. x7(t)in Figure 17.g. however. procliacrs the signals. for a receiver. 1. it is pointless already krrowii to the rcwiver.oinrnunications. the greatrr t h i ~ ~ f ~ r i con4. Jn fact.mdorn signals in tlectroiiict. The first amplifier would aiso never repeat the noise 4gnaI x ~ ( t )so we cnn do very little . IVe can. Examples of ..tic property which can br cdnilat. Iiidividiial rantioixr signals (e. A process that prodrtces random sign& will be (died a randoin pmctss.We will concentrate ~ OKI 1 he rmdorn process that prodncrs t k signals as only it giws iiifoi~~~iatiO~i 0x1 all its sample fimctions..1) a r called sample fur ons or retzlrsatzons of a random p~ocess.1 mid 17. Thc solution to this problc~iiis found not? by considering individual random signals. the less that can bc predicted. Thcs kntm&dge of this output sigrial. This h w two Ggnifisaiit admritages: the given noise power is a determinis. x l ( t ) . noise.sdf caiilrnot be iiiathematically dcscribed. it.2. in Figurcs 17. for example. hut uuefril signals c a i ~ also he stoch 1x1 c. c'anitot he transtcrrrd to other sitnations. iiiterprrt the signal in Figiirc 17. for exampie. are ~ ~ ~ t ~ ~ rsignals like (antenna. hut instead by ttnalysiiig the procew that. it raii still be mexsiir~d and a graph can be nbtaiiied. . TOdescribe rantiorri signals we cnn first of all try to Start wirh tlir signals theinstIlvcs. liistcad. It is mt known whether Fourier. The emttixety of all random signals that it. We need to introdiicc some nrw terms to extexid our discussion to general random signals. Drscribing Ilaridoni Signals Sigiials that ltevr unknown behavioin arc called rwtdeter m 1 r 1 1 . . we should derive general st#alemeutsabottt the noire bc. can prodixccb is caIIed an rnsernble of vandoni signds. the r + sulting . Of coursc. the typical noise ponver can be given.
. which are found by averagirig one sample function along the tinieaxis. . finite niimber of sample ftnwtions can give an approximation of tlie expected value from (17.rl(t).c: ensemble of random signals (expected value). 1 eete lWS 17. we use statzstzcal averages a. and tzmeaverages.h expected values and timeavera.2) In Figure .1 Expected Value and Ensemble Mean Several different sarnplc functions of a process arc reprcseiited in Figure 17. time t l . corriplet.1. We need tools from mathematic*alprobability. Expected Values 405 To characterise randoni signals deterininistically.(t) takcs anothcr value at time can bp observed that the meail or the fimctions 1 2 than at. In the next section we will deal wit. not just srlectcd sample fiinctions.1) being only part i dly carried out. there are three wailable methods.i an avrrage a(~05. They c m be classificd into characteristics of random processes which hold for a. e h e r a g i n g a. Thc approxirnation becomes mor e acciu ate as more sample lunctiorls are included. .ges. that are measnred at the same time on various amplifiers of the same modcl. As expected value (also ensemblr mean) we define the nieaxi value that is trt)tained at thc same time from all sample functions of the same process: (17.2. The definition of the expected value in (17 1) should be understood ac.17. that. tlie tirneaverage i s taken in the direction of the timraxis. . the expected valiie i. .ra1 timedgycaderit : (17. different times. The expected valiie identifies the whole process.lso called just averages. however.1) A s v.s pro the .1). M'e can imagine that they are noise signals. the expected value is in gent. Sincc the averaging in Figure 17. we did not want as prerequisites. In contrast. This is equivalent to the liniit in (17. a tormal tiescription aud not as a tiictliotl for its cal ion. It says that it shonld be deterrninrd from all sample functions of a pr which is in practiccl an impossible task.2 2 ( t ) .c. If we wish to dttiially cvalute an expected value. .e can obtairr differeiit means at. 17.2. and is an ave~age ulanq the process.1 runs in the direction of the tladiecl linrs.1 it * from precisc knowledge of the process the cxpected valiie can be calculaled without averaging sample functions.
1 We use the example of a die to show the three methods for finding an exact ox approximate expected value: 6 Fkom howledge of the process. e the same for If wc repeat the procedure with ~ n a n ydice whose syrnrnetxy properties we do riot know precisely. Example 17. Describing Random Signals I I 0 e I I Figure 17. the sairie time.406 I 17. so thc expectcd v:tlur is: E { z ( t ) }= 142 + 3 + 4 + 5 + 6 = 3.5 * 6 is As the die's syrrirrietry does riot change.1: Example of sample functiurls 2 .( t ) 6 Under certain preconditions. wc can also approximately calculate the eiisenible average by averaging all of the iiurrhm on tlie die's faces. we should take the ensemble Iiiean for all dice at. . we can say that every number has the same prob. the enscnible average can also be expressed by the timeaverage for a sample function. that we will deal with later in niorc dctail. which is the complete symirictry of the die. As the expected values can in general change over time. the expected valiie any time t.thility of appearing.
2.3) for f(.1). The mean according to (17. Expected Values 407 * Changing statistical properties over time is not a.1) is contained in (17. we will discuss sonic more general forms of expected valiies.3) In contrast to (17. The expected value E { z ( t ) )tells us what value to expect on average from a random process. We can further say that the average for niariy dice at the snme time will yield approxirnately the same results as one die thrown repeatedly. Before proceeding with the relationships betwccn the ensembleavexage arid tirneaverage. z(L) is here replaced by a function J(n:(t)).average. where Glie values at. come up are the discrete values of a sample function along the time axis. In order to describe such properties we introduce the general .3) for f(x) = 2.r) = (z .verage. the noise power of an amplifier witli~~ut lo. In this ease.p . but it does not fully cliaracterist. ~~ sample functions of random process B vary much more around the axerage than those of process A. different firstorder averages are obtained. For f i x ) = x2 we obtain the ~ ~ a ~ u~ ~u ~~ .17. the t.imea. It is obtained fronr (17. The squaw of the deviation from the fillear average is also important. We will soon feasn about higherorder expected values.~but ~ ~ e ~ v ~ r ~ they clearly differ in otlier € ) ~ ~ ) ~ e rThe e s . for example.2 shows sample ftmclioiis of two random processes that have the same ( t ~ ~ ~ e . The numbers that. The reason why we use the term firstorder expected values is because they oniy take into account the a ~ ~ ~ i t uof the sample f u n c t ~ at ~ ~ s des ~ one point in time.?xl. By choosing different fnnctions f . Figure 17.firstordcr expecled vulue: (17.5) . It is also cdled the Eineur avemge and is denoted by E(z(t)}= pz(t). the process. s ) 2and is called the uarmncc: (17.4) M’c can use it to describe the average power of a random process. g e :c 7 e ~ z (17. concern for normal dice.more than one time are combined together. the expected value can also be expressed 135.
2 ~ ~ l A general characterisation with firsbordcr expected values is obtained by pntting j ( z ) = E ( @ ) . The rttlldo~xr process A in Figure 17.408 17. the standard deviutzorr a z ( t ) .3) and determining tbr expected valuc dependent 011 tht. We . When combined with secondorder exp sufficknt to chara rise many common random signals. Variitricte and standard deviation are measures for the spread of the amplitudr around the linear averitge for a randnnl signal. E x ~ 17.2 deariy has a lower variarice tlrrari random pro( The linear average arid the variance are by far the most frequently used firstordcr pxprctcd values. Describing Rmdorri Signals random process A random process B Figuse 17. For general raridorn signals it is possible to define inore firstorder expected wliicv by choosing a different f ( x ) in (17 3).2: Randoiri processes A and I with a diffwirig distribution between thc indl3 viduitl smiple furletions Taking the positive square root of the variancr giwt.. into (17. threshold 0.
17.3: Distribution frrrrrtion for tlie Iiurribers on a die For the die fruiri Figure 17. Thc rlrrivative fi’igtirc~ 17.3 for the die from Exninple 17.) i s a liiicar operatotor for which the principle of sixpergosition holds IIere ( I and B can also be deterministic functions of time u ( t ) and h ( t ) .2. if thc eorrcspunding calculation rtxlcs ale observe& There are actually only two simple rules to remeinbcr : The expcrtcct value E(.3.1. It is clrpictcd in Fignro 17. Expected valiirs 409 It rebums t. Expected values can Itet dcalt with in a similar way to frtnctions.hc>probability that r ( t )is less than the thrediold 8. as in Example 17. this yields a h e a r amrage pz(l) = 3. .5.1.
9) for f ( p .xpecr.2p2(t)E(~(t)} p Z 2 ( t ) + = E{x2(5)} . which is obtained from (17.rexnem: 0. s i ( t ) and zf(t)are summed and the rcsult.p:c(t))Z}= E(z‘((1.10) It describes the relationship between the whirs of a raurdom signal at t i m e s tl a n d t 2 .).7) Example 17.ic siglial d ( t ) is the value of the signal itself. I (17.a i d then the third can be calculated.ity z i ( t ) . and tliereforr tlrcay cannot register the statistical dependencies that exist.ioiisare available. is divided by N . With secondorder expected values.8). v) . Firstorder expcctcd values hold for a certain point in tirnc.2z(t)p. For (1.1 = C2.11) .410 17. A’ The autocoxrelation function (ACJ. a5 d ( t ) ran be thought of as a realisation of a random process with identical saniple furictions: I E{d(t)} = d(t).{x2(t)}. Jkscribiug Rmdorn Signals Tlre expected value of a dtltcrminist. ‘I‘hcy link the signal at two different points: .3 We use both rules (17.~d value.ilable. High values for the autocorrelation function indicate that ~ ( t )times f 1 at.p i / : 1 mean square IPrz(t1. this is possible.L2ie linear average.Q ( t ) = (17. the autocorrelation functioii becomes the I P u s ( 5 1 . the square average and the twiance .p z ( t ) cannot be averargedbecause iii the first. 2 ( t ) = E{(s(C) . If N saniple filnct. ava. t l ) = E{r2(tl))‘ (17. hetween diffcwnt points in c2 signal. the linear average p c ( t )is rmt yet. however. The qu. t 2 ) = E{Z:(Ll) t 2 ) ) . p a s .7) to express the varia. The relationship between them is often iised in practice tro calculat. 4 1 (17.(t) pzj2(t)} + E.’) is an important secondordrr c.6) a n d (17. and ta take similar values.tnt. A secaiid pass can bc avoided by calculating the vmiance in accordance with (17.8) It is therefore suficierrt to know only two of the three quantities .nce witli the linear average and squa.e Ihe xuiance.
stnttstacs. % l random processes A arid B hmr identical first‘w order expected vatraes. This i s confirmed hy the measured ACFs shown in Figure 17. This ernergiiig branch of signal analysis i s therefore called lazqharorder. In Section 172. We can therefore expect a much greater t.t z > for 13. Stationarv H mdom Processes 41 I Example 17.3.4 shows .1 we considered various possible methods for calculatiiig expected One of i h s e w a b expressing the ensembte inem with the timemerage. rand^^ process A rand~m process E 3 . The conditions tinder which this is possible will be explaitird in this section.4: Illustration of two raiidain processes A4and XJ with iderititat Emtorder expctt. However. fmction o y ~ values tl and 12 than for the ~ ~ ~ t o . in particular their distribution function Fl.2). * t t Figure 17.tvliicli pmperties ol a rarirlorri signal clre represenl ed by secondorder expected values.!ti (@) (Example 17.5.aliir of autocorrelation firriction pZ5 Cz) for A’s neighbouring (tl.17. Many descriptions of stocbastic funrtioiis rely d e l y on these two. Tlie sample frxiictions ttf random prnrcsh A chmgp mach morr slowly with timc. WI~LE~. tlrtzri thosc of process B. . advauced models of complicated random processes resort! a140 to higherorder expccted values.aiid secondorder are by Iar the niosl important in practical applic‘ations. however.c ~ ) r r e ~ a ~ iq y~ ( f l .4 Figure 17.cd values ancl differing becolidorcler expected values Expected valucs of first.jt) (0)= Pf.
as i?. for which the linear cbxpec. If it is lorrrird froni a signal whose statistical properties do not chaxigc with tirne.2.41 2  17.!3). We c a i itse fliih pxoperty to titifine ___ tl a i md 12.tetl value is eqial to the current functrrm value.)) a special case of se<mdordt. This means that dctwninistic signals can only be stationary if they are rotistant. signal tlial i s z ( m before or after H ical properties and cannot be stationary. ) .12.  In (17.f I f s secondorder ~rpecttduulws only depend of ohswved tinre poznts z = tl . billne amount At: The expccted value does not depend ox1 the individual time points irtstcacl on thcir difference. for tiirite random signals. r ( t . For a precis(> definition we start with a secondordrr cxpccted \due as in (17. in nrinistic.r i~xpecl are wlues ~ 7 find: e . it is clew that tlic linear expected value changes with tirnc.4.. certain point in time changes its s From this cicfinition for ~ C C C d valurs we can derive t l r c properties of fiist omder expected v d l ( . but 23: Stationary tdom pi*oeesc as stationary z.7) we cvtisidcred dctcrininistic signals as a spccial case of random signals. if both time points tj nut1 t 2 are shifted by the. wliilc in Figure 17. then t h e expected value does i i o l charigt. signal that changes with tirnc is thcrr)fore not statiorrery. This appwrs to be the case for 11ie two random processes in Figure 17. Describing Naiidorii Sigiials  A ranclom process is callcd statoonnry if its statistical properties do miot change with tinit.r(t2))== g(r(t.
when only the linear average and auLtw.which is a funcliori of two vttriables.c ~ ~ r ~ I a . tmd not for gewral funcLions f ( . u L o .).) . ~ ~ i ~ e fuiiction is not the~ n ~ ~ o ~ a same fimctitm as pzz(tl.ls R oriefunction z). ~ i ~ n pZ2( that we introduced in (17.lic given ~ I ~ ~ o r ~nniiottellt whtxthcr or not the r a n d ~ ~ n tve ~ ~ ~ ~ o ~ process is stationary in tlic strict mnse . function of weak stationary proc PS (17.riance: /&(tj = ps. If the stationary coiidition only holds for (17. weal<statioiiary is n less limit in$ condition that that of (strictly~ stat ionary proceqscs for all expected valiics of first. 0 2 2 ( 1 ) = o.15) The a .131.and swordorder.12) is fulfilled for the funchxis (17.tz: E{Z(tl)s(t2)) f= E(zftl)z(tl. Stationary Handoni Processes 413 That means that for s t a t i o ~ random processes.15j .7). prclcondition. = E { 2 > = V L I 0 ) = ( I. We can thus obttiin the following staternent: for a weak stationary process the linear average and the correlatioii properties coiitainecl in the autocorrclation function are is iisually used in constant with time. The two pE1: linked for sta.l.101.. t L and t z . the vaxiaace iilid the ~ ~ u ~ o . .15) hold.17. (17. In partictilar. Thus (17.c')) = E { ~ ( t + ~ > ~ : f l : z a )pXz(.orrelation fund ion itre coxisidered. } (17.15).17) however.5 M e take a random piocess with linear average E{ t ( f ) } 7 0 and the ACF pzz(tl. where it is often a.~ o r r e i a ~ i o ~ Also ~ the ~ r avex . (17. Example 17. Thir r n ~ ~ ~ ~ ( ~ ris definitely weakly ~ t ~ t i o n n r ~ .62).14). the firstorder expected wfiies ~r~ do not ctepeitd oil time. z * (17. The vaiiuncv of the signal ir cr.+riditi linear merage p z = 0 i s constant. = sijtl t.16) (17..16).).tz) in (17.3. for the linear average and the va. . can mite its AC'F ah proress n as we i F r z ( d = "i( ) . With t.14) The autocorrelation filnction c m be expressed more simply with z = tl . The concept of weak stationary proces the context of rnotlelling and analysis of random procesms.tionaxyrandom are sigmls (17. arid (17. then the random process ts called weak s ~ ~ ~ ~ ~ for ~ ( linear ~ 1 age.
sin(wt)c(t). = ?72(tl)m(t2)piI ( t l . It.18) cud the secondorder time average (17. it is WC will now return t o the qurstion of the conditions under which wr ci>n esyrcss the rnsenible averagcs by the time average. For cxarnple..414 17. this is riot true for nz(tl)in(t2) sirid1 sinwtz.5).5 i s modulated by a deterministic signal r n ( f ) = sin ~ v tso a new randnrii signal . the mean square at Liine t = & is but at lime C = 0.1'3) . W e h s t define t.6 The weak stationay random signal x ( t ) from Example 1'7. = Therefore y ( t ) is neither st&ioriaiy nor weak stationary. y(tj = m(tj. ti').r(t). t z ) only deperids 011 thr differeiicc between the observativri times z = t~tz (see F:xaryIiple 17. but.liP firstorder time (17. p . 13escribing Randoiii Signals Example 17.y(t)is created: .
The big advantage of ergodic processes i s that knowledge of ari individiial saniplr function is sufficient to r. All phase angles are equally likely. .v. The prates i s statiomry because the secondorder expected wlues (17. (I ngree for all sample functions of a raiidoni process and tire also ecpal to t lie idle average. Example 17. we can then express it with the time average of any sample fu . If the ergotlicity conditions orily liold for but not for general fimctions . does iiot exist tor a ~ oo ~ f oi i).3culate expected valiies with the timeaverage. as long as no indications to the contrary occur.12) oiily depend 011 the differencc betwcen the observation times. It must Ite prown for iridividual eases wfiether or nor a s t ~ t ~ o nra~ ~ ~ ~ d ( ~ i r ~ 1 ~ pro is ergodic. but the phase pzis couipletely random. Similar to stationary processes. thexe i s also R iestricciori for certain random RCS.3. Often this proof rannot be exactly carried out. Stationary Raridtrrri Processes The ~pperentlycumbersome limit T CG 41 5 oc IS 4 nec ry because thc iih rgral o f f ( . arid in these cases it tan be nssiinicd that the process is ergodic. f i e . The ~ t nar~ ~ ~ i average is thrrefttre fomicd from a section o the signal with lmgth 2T a d this f bngth is then extended to irifiiiity.). &heraiitbm prnrfsi: is then weak rryod~.18). ~ ~ ~ ~ i o random bignal. Like the idea of weak stationary. I__ l l l _ ~~~~~~~o~~ Ergodic 24: A s ~ ~ t mnilom ~ wss ~ n which~the &menvcragcs of each Jar ~ ~ ? lor ~ the sunre as the enscntble uvcmqe LT talled an ergoiiic raii~fomprocess. ) c t ~ . for example.7 A random proccss produces sample functions wkiere L J ~ is fixcd yuarAiL. Random pio~essrs this kind are called of eryodzc. It the tirrieav(irages (17. it is used €or modelling arid analysis of random proccwcs with miniilia1 rcstrictioii. The ACE'.17.
imeavera~ge tlic square of a cert. the expected values c m be with time avrragc~s.416 17.and secondordei expected W ~ U P S .cosfw0tl w&) 2 = . We want to describe t.i a t the bcginiiing of this chitpter. tlw autocon.he propel ties of random signals with deterministic. We can verify this with the rxample of the ACT. W e start with ergodic random processrs if needed.d valiie e the linear avm age. we van return to the. . Describing Randoiii Signals 1 = .cosw(Jz.19) agrees with the ensemble average. The process i s also crgodic because the timeaverage (17.ain sample function i is. vaiiables* This will IWW be doric with first.22).t 2 I The integral averages all phase angles between 0 arid 2n. brit ~ = l ~ Now that we are comtort&le with expected values and Iiavo learnt ways of c a l m lclting t h i n . zP(t) = of it i s indeed statiotrary. however.xpectt. so that. 2 If the random process tdso has am raadom peak vdue P in addition to the random phase p7so that ~ ( t )2 sin(uoO% 'p>I = t not ergoclic. For cxanlpJer while E{z2(t)} iE{.rlation f~iizctioiiand a generalisation derived froin it.coswgI z mit 1 2 z = tl . where 1 __ . The most irriportant c. the t. task from Section 17. tlre crosscorrelatiuu function.
4. The restllts will latcr bc cxtencled to toinplex signals. We will get to knoa the other expected values better in the lollowing sections. iu particaular t irv autocorrelatioiiand croshcorrclation fiirrctions.20).22)~ rio 1c)iige~ f l f ( ~ is riidrkivf hy an iurkx (cnirrptiw (17. the st~mplc i ~ ~ r iii .14) arid (17. AS LEE time average c m in t31iiscase kc iorincd from a i y sample t fimrtittn.18)}. From Defiriil ion 24 it is ecluill 10 the As ergodir functions arc sttitiwtiry.18) vie ohtaiii T (17. in accorilarrce with (17.lk crgodicitv i d 1 br. As me only use secondorder rxpectetl vaiizes LPL the f o r m o (17. we.22) i s (iikc all other first order expected valires) indeperitleiil of time. Correlation Funrtioiis 4 17 For the linear expected value. A f sufficiciit pic>rccluisite.17. . the finrar average p time average m. (17. First WP cwnsider c o l d a t ion functions of purely red signals.( ~7.lf.
. the displacemciit between the z) two fiinctioiis in the product :c( (t r) we can subvt.ute = t z and this yields + %’ + .it.29) (17. J) There is 110 sltift. between ~ ( t ) x ( t t.(t’)}= E { X ( t ’ ) l ( t ’ . 110 general i statemrnts can be made. {X(t)Xx:(t E + T)} = E {s(t’ . t h e is no relationship between distantly separated v~tlties.33) For signals tvlrose values are uncorrelatd if Pm apai t .~(0)21. As (. thrc.z) for which the expected valiie and f ~ c o ~ iE { xs t ) x ( I.32) E { s ( t ) }= E { x ( f .6 shows a typical autocorrelation function with the properties just tliscixssetl: .in nc:cordance with (17. If the shift by z is exwtdy equal to a shift. For the Lehdviour of the autocorrelation function for r .. . to ZT = 0.he value of pz!r( only depends on. by one period or a nirxltiple. L)cscribiiig Raiidorii Signals If we use (17. +. this Tnea. and therefore f‘or the airtocorrc?latioiifunt.418 17.z)} = fL.z)} (17.. such that thr serondorder expected value is decomposed into the product of two firstordei expected values: E {J’(t)Z(C  z)} = E{+)} * E {r(t  T)} jz*/ +OG 011 (17.25) on c p g g ( r ) : t. The value pzz(0) can also he expressed by the variance C T ~nix1 tthc + lirieas mean pT.he ACF of the signal is prCT( 2 cp!i.~i:(O) i p: = z) ~p. however..z)x. 2 5 vm(4 5 cpZS(0) = 0. (17.ns that the preconditions for a. 2 ‘ (17.ttion fuiiction is it.1.30) is then iaimediately obtained. (17.c(z) = r19 IT1 + 33. the equals sigii represents the case where ~ ( tis) a periodic function of time..31) As the linear expected value of a.r )1 > E ( z 2( t ) } . the only relationship between t hmn i s the (timeindependent) linear average p Figure 17. This property is expressed in the expect7edvaliies of the signnl.These values are then said to be uncorrelafed.ocorrel. sginnietry with respect. si atinnwy signal does not depend time.tion ip.4): Vm(O) + 2pz = 4 + p.CO. In maiiy cases. weak stationaxy process are not.28) In the relatioiisliip (pZz( r ) 5 cpz:c (O). f such an efIect! i s observed when measnring i~ ( X an ACF. A further property of the aut. then s(t)z(t c) = x 2 ( t )and also ( P ~ : ~ : ( Z= yzZ(0).
~ ( O ) ~ ~ iis reyuill to value ~ ~ itself i . ) ) The antocorrelaliorr function pJ ( T) in particiilar i. in Figurc 17. The r ~ ~ ~ ~ ~ i P .5. but. 'Yhe dgnificantlt. very sirnilar to the autocorrelation function fIomii Fignrc 17.37) only holds when ry distantly separ2zted values do rot correlate.7.1) ~ ~ ~ r (17 35) > i (17. v ( t l . Under the assumpi 1011 tliat these processes are also stationdry for the sectiorr shown. the Imear average i s zero.4. t ~and p z . 11: ~ pzz{z) (17. we can charace t h a n with their autocoir ion function:. q z 2 P) ailcl pvv( shown in ( t). along the I I or 1 2 aAis.36) hold for all ~ t ~ t i o i i a ranciom signals t ( l ) .3.7. faster changes in the t ~ ~ i ~ ~ e ~ of: v i ( ~ i process f3 in Figurc 17.8 Iii Examplc 17. which i s also visible from the sample furwtionc.6. t lin Figure 17. $. (17. Correlathi Fimctions 41 9 i ~ ~ value: ~ ~ pZ2(~1)~ ~ = pJz( z) i. Example 17.17. In this casc. In Examplc 17.4 a rnncEoiii ~ r are also expressctf by i~ fastei deray of the a u t ~ ~ c ~ ~ ~ r l cfunction p q uz) from atiori J its maxiiniim valrie in Figiiic 17.37) does not hold.liose saxxi( p l ~ fiinctions c*lemly have statistical interdepeitdencies. They ale illustrated in Figure 17. mid ale tountl hiiiiplv as riowsrctionv throngh c p r z ( f i .4.. for example.7.4 we corisidered two ranclom processes A m d I vc.p ~' Itmer bound: 2 0. Figure 17 7. (17. lio\vever.
l xaaxple 17.random process A random process €3 FG J> ..9 .
The properties of the airtorovariance function coliespor~dto those of the autowrrelation function for z e ~ o mean signals. it In order to o'vercorrie I b e difficulty statcd a l m ~ e .4. This idea can lie extcndcd to signal value:. but the second random process i s irrdimtcd by another letter in the indcx.correluhon finct/on.2.hat (comparc (17. the linear average p.10)): (1 7. from different rsncloni pr cnt i t s propertics expecled vitluc is called tlie crosr. A sccondoidergoznt exper Led valur~ the expectd valiir of a function j ( ~ ( i ' ~ ) . the linear average can be rernovtitl horn thc outsct and instead of the signal . ~ ( r): ui/?.P: 3 (17. .PTl(.17.the zero mean srgrd ( r ( f . is CL)).2).. t h i s is suficieiit condition for the c3xistcnce of the cleterministic functions. it holds in general !. Correlatioii Fimctions 423 time can be absolutely integrated (see Chapter 9 2.4.i denoted like the autocorrelation limrtion. / ( t l . correctly we have to extencl the earlicr defintions of se.3 WP obtain ( 7 )= P'4 J( .~ 1 % )can be cousidereti.i that arc take11 from 07ic Iandorri proccss at LIYO tliffercnt times. is non~ero. Its aiito correlation function is called t l i ~ ) autocovurmrm functzon of r ( l ) and is drnoktl by V J .V(l  z p5)).1.hi { d t l ) '  "!"U The rf(i~se0rr~7131ioii ftinction i.c(t>. formed with sigiials from two diffeient raztdoiii proce (17.39) just as in (17 8). This property is not giveri for autocorrelation fimctions.cmdstatiomry and ergodic rnndoni processes to ded with two random pzoccsses. ) (17.38) IJsing the calculation rules from Section 17.45) i p l y G i l l z ! =.(T) = E{(z(t) .3 CrossCorrelation Function The ailtocorrelation funrtion is given by tlie expected value of two sigrral va1ue. t z ) .44) For the crosscorrelation function y. how~ver. 17.
~~~2)} = ?dfz). There ?axe also weak forms for joint. Thc crosscorrelation function performs it sinrilar fiirictiori for two random processes Illat the autocorrelation ftinrtion does for one random process. Describing.50). two rimtforn pror. s. ~the rcros4 corxelatiori function clors not have the even s ~ ~ ~ ~ e e ~ n ~ ~ ~ ~ ~ of the autocmreiation functioa.(t1)a d ~ ~ ~ = ~ ~ ~ =) { ( ~ ~ . ~~ { ( ~ ~ ~ ~ ~ 3. we cmt find Ihe syiiimotry propprty (1736) of the autocorrelation function. the of protlurt of the linear cxpcc%Xl valuos 11~and puy the individual random processts: of Y.46) Finally WP introduce the secondordw joznt tzrneaverage and call two random processes for which thr @hit cq)cct. older exxprctccl V ~ L I only depend on tlie difference z = tl . wv omly obtain The ai~tororrelat. .~~. ( f t Z). are only fulfilleci for ~ ~ ~ ( ~ ~ ) ~.yft))=:E{z(t)yid. It is a measure for the rF~atio~sliip valnes frorn the t~ 7 o a r i ~ o praccss at two timcs of r i~ separated by z.z) ' ) (17. The extension to two random processes cat1 some (jiffererices t o the antocorrelatioii function.it). as from (17.15): c p & ) = E ( . x . brit at least for 174 30:  F t ~ r ~ .~(~~~. For joint station~ S ary random processes the crosscorrrlntion function then takes a form simifar to (17.ed valws agree with thc joint timwiverages joznt crgodzc. Then using (17. stai Liona.48) values of TIicre i h also the caiw that two riuidorn processes art? riot uxit:orielated for all r.46) and scvagping n: and g. Their crosscorrelation fiinctioii is then. hndoxn Signals Next wc extend the idea of stationary from Defiirition 23 t o cover two randonr processes.ioir function pTJ z) can be obtained from tlie cross( correktion €ianc+ion pfy(7)% a special case y ( t ) = . We call two random procescies jornt ~ ~ u ~ if z o joint~ secondtheir ~ ~ . (17.ry atid joint crgodic randovn p r o c ~ s s ~whcre tlic corresponding conditions.es~s be uncorrefateti not only for large timecan spaiis hilt also for all ~Tlfues r.&j = i*L Ply v r. Firs! of till.tz.422 17.
p g ) slid this leads to the crosscouanance fur~rfto.2) In this section we will therefore be extending the use of correlation functions to complex random processes. These are rmdoxn processes that produce r*omplcx sample fiinctions. When 'cve ~ r i t r ~ ) c ~ the evarious correlation functions in Scction 11.2.4.4Ci).39).( z) and crosswrrriation function plw(t).n (compare Section 17.2): (17.4. Correlaliori Functions f "14.1 ~ ~ 0 ~ ~ .51) li/zy( 4 = W j 4 f f .17. the calcnlatioii rules froin Section 17.P y H * As with tlie autocovariancc fuiiction in (17. "e define the c~osscor~elat.C o r ~ ~ ~ a t i ~ ~ Function Thcrc are several possibilitics for extending the crosscorrel.i~)n fuuctioti for corngkx random pr webses as (17. autocorrelmtion. . For real random processes (17.ion function and introduccd the crosscc~rrelatioii fuiiction as a generalization that conhined other cox relat ioii fririctions (crosscovaiiance. c o r r c ~ afunction for roniptex signials arid derive tio~~ the othei correlation furictioiis from it.4. aut owvariance) as spc'e h 1 ca5es.1. however. complex sigrials will appear as in tlie sigiisl transmission example (15.46). According to (17.4 . 'Fo introduce the correlation fuiictions foi torrrpkex signals we proceed differently to Secliori 17.ions.3 yield tlie relationship between crosscc)vmiance function 7$zc. 17.1 ~c d to rttal signals for the sake of simplicity. SIere we start w i t h the c ~ ~ s s . There are differ mt definitionh in other books (for example 1191).J.53) The only difftw.1.rel7ce to tlie definition for real random g)rocesses is that the conjugate complex ~~11ctioii of tiinc ~ ' ( 6 ) is used.C ~ v ai ~i i~a t~ i~~~n~ ~ The c*rosscorrelationfuiiction can also be forrned foi aeromean signals ( ~ ( .4.P T M t . ' IP C' will clioose a definition that allows a particularly straight.53) becomes (17.p7) t) and ( p i t ) .~tinr1 function to ccnw complex random processes.2.4. forward intcqwetation of lhe crofispotver spectrum. There 'cve started with the aixtocorrelal. 4 II 423 C r 0 ~ ~ . In many applicat. To do this we niust ~tbsu~rie a ( t ) and that y(l) represent complex random processes Ihai are joint weak stationary.
ll odd imagiiiaxy part of pLZCf For real random yrtrtesstxs r). With I h e same reasoning as in Swtinri 17 4. tlie a ~ t o . . pS5(0) is also purely real.53). = )) / + (17. for y ( t f = rc(t): As in Ihe red case. We cmi obtain it dirrctly from (1’7.57) I or inorct concisely (€7. the crosscorrelation function for complex raxtdonr processes i s also neither symmetrical nor comrtiutative: For uncorrelated rmdoin process we obtain 17.). .1.2.z)nqt’>}l4 = p (.2 Auto.squsiecl.*(”)]*) = .) = [I? { r ( t. so for coinplex random processcs.4. = 04 l p r 12 .and is expressed as r w n rcid part and a. ( I 7. Describing Rnridoin Signals In geiicral.58) The corijugitte syrnrnelry here van be recognised from (9. and p7I i ~ ~ ~ i t can be exprcssed by the variance and the inem j c T : itlt I ~2 pr2(r) 5 p&%(O)E { .56) by substilir(irig t’ = t i r and by uhiiig tlir c~alcuiattiom rules for conjugafr cornplw quxitit ies: pf&L(Z) = E ( z ( tf r)*c*(t)) E{T(tf). the imtcgiiiary p x t of p k i ( T) i s 7 c and tho cvcn symrnctry holds accordiiig to ~ (17..36).424 17. tltc J&”hilcthe ineau variarm is atways a rcal qnant ity as tlie squzire expected valw is formed in this cane with t h e magnitrxilc.4i)).Correlation ~ ~ ~ t i o ~ The autocorrelation fuuvtion of a complex raritfom proccss can be obbiried from the crosscorrrlathn furlclion as in (17. c ( f ) T @ ( i= CT: 4pTp.59) is in general eomplcx for a complex random pmctw. h aiiy case.E .r) .c*(t’ z) = E ( [ x ( l f.c o r r ~ ~ l fmctiou icolisifits of a symmetry relaa~i(~~ tion at the transiliori fiotn r to .1 it bolds that the ~ i i a ~of ~ ~ ( r ) its rnaxirnd at z = 0 .. the odd iruaginary part tlivnppears at T = 0.
62) er Describing ramloni sigrials with exptctcd vdiscs aiid in particular wit IT correlation firrtc'2 i o r tdirwed that the properties of' rwritloin sigsials can ktc cxpressed by deter~ rrrinistic ~ ~ i a ~ t ~ ~ ~ raritiorn t iinc signal there would f w ~ cximiple. qiinntity.17. for thci [ l ( ~ t ~ 3 r ~ i ~ i s t i c atttororrclatioi function which is tiktwiw r2 tletermiuistic. Power Densitv SDectrii 425 (17.5. In the past chap we saw that describing signals in the freyuertc)rcloinain is . Iristead of a ~ s .
but f this idea is actually unsuitable.ed at~solntely j9. We start with tlie autocorrelation function or t8hem.69) It B olso C R I I C ~ thc poituer denszty s ~ ~ c t of‘7 6 random prcicess. i c )r~ ain quantities to the freqwricydomain.~ ~ ~and ithen transfer blre ~ e ~ e r ~ i i ~ i ~ s t . = 0 . as they do not decay for ltj + 00. Instead of transforming into the frequenc.ydoi~iaju and then fornrhg rxp fwirt the expcrtcd values iK1 the t i ~ ~ i e . ~ o r ~ ~ s ~ ? o i i d i n Fourier tr tmhrxn of a crossthe ~ ~ y .{X(p) 1.4).426 17. We would therefore like to have a detemiinistic dcscription of raxidoni sig~ials in the frequcncydomain izs well Thc idea that Erst comes to mind i s describing sairiplefianction doru process as expected va1ut. Dcscribing Rarrciorn Signals part itularly uscful. correlation fimt%ion can Be fornied giving t hc cyasspowcr ilertszty sprwYurrt: Its h e a r average F. Stat ionmy random signals can never be i&egrat. This idea is the hack fox the dehition of thc power density spectriirn. The power ~ the ~ ~ density spec(ruin chararter ises statistiml dcpcndcricics of the signal amplitude at two diffcvent poiiits in time. Therefore the Laplace integral cmirot exist a i d thc Fouricr transform earl only exist in qpecial cases.&ocovariancefunction of a weak stationary random process arid form its Pourier transform: (17. art1 often Ic5.s instead o considering them as random signals.zds to elegant aieL1iocls for aualysing LTXsystems.
we can obtain the relalioiiship between the power tfeiisity spectrum arid mean square by putting z = 0 into (17.5.ti.) pg:r ( t)= .:.0 0) = 7r 2 4 qw 2 IT +U{. multiplied by a factor 1/2n.ly from the power density spectrum. It i s similar to E:{IX. In the best case we ctm form 427 as a deterministic desc*riptionof the rancloiii process.71): z‘ The iiiean square is therefore equal to the integral of the power deusity sprctruni.(p)I} and likewise indimtm that the random signal oiily contailis frccpeiicy tompoiimts at &WO. Power Derrsity Spectra cannot be given because of the delta impulse.  The mean square of a xandorii process c m also be calcrrlated direct.um proportional to thr meati square of the crignnl @.(JOJ) or a signal r ( t ) is . the auttrwrrelstion function is expressed as tlic i n ~ ~ r Fourier transform of the power density spec%rum: se A s the mean square is equal to the value of the autocorrelation fimctiou at == 0.9: Thc arm uadcr the power dcusity q)ec. area is equal to the mean square of the signal (“power“) Figuie 17.. First.:(jO) 6(0 . Forming t hc cxpected values first in thc timedomain and then transforining thein yield? the power densitv spect r i m @:z.cos U ( )t 2 I as the Fourier transform of the ACF.17.
Drscribiiig Iittiidom Signals .428 17.
5.79) This means that. The toncent. bite Noise Many iiiterfcrence sources. cc2n be described by rmdoni processes with a power density spectrum that is almost curistant over a Iltrge mige of frrquencies. like amplifier noise or radio iiiterferente. resnlt agreeh . of courbe. which represpnts B uniform mixture of.ra 429 Example 17. In Figure 17. the idealised process is called wh7te nose.ration of power density in t. . it is real and positive.9. samples of white noise taken at diEerent limes do not correlatr. Power Density Specl. Such random processes are ofteii approximated by an idealised process. ( (17.viiith the rough ~ s t ~ i ~ ~ i oft ittir ~ n ~ ~ c n t a 1 This a o~i speech frequenry in Example 17 9 Figrtrc 17.10 therefore. T h e ~ ~ i ~ ~ c o r r €unction ~ ~ ~ e l a t i of pihike noise can only be r ~ ~ r e s e n byeit dis~ c~ tribution. every coluiir in tine spectrum. itself rpi&valucd.he 100 Hz region is clearly rec(~~~iiseab1~. only the positive frequencies are shown.11 Ab 8 1 example we will consider the power drnsity spectrtxin in Figrrre 17. the power density spectrum is also an even function. the t3clta impulse: (Pr.5. T h e name is dcrivcd finin the term 'white light'.&* ( r )L: F"(NbI =r No S z) .10 29. whose pomw density spectrum i s not freyutwy dependent: Since all frequtmcies occur evenly. anti al as the speech signal is.17. As expected.111: Paww density spcctriirn of the speech signal from Figure 17.10 x for tlie speech ~ i g ~ ifrom Example 27.
hc timedomain (17.ot actually be realised. a white iioise sigad niust have infinite p0WeT: (17. correIation function. characterises noise processes whose p0~7eris evcrrly distributed below a band limit* w ~ ~ Figure 17.11 stiows the power density sItectrxm and the auto* ~ ~ . A s most of the r ~ ? ~ s oand ~ ~ ~ ~ i v a ~very n ni ~ is .11: Bandlimited white noiw in the frrqwncyclomain and timedomain The concepts discussed so fw for corittinuous rarzdom signals can c~tstsily extmdod be to random seqnences.nixo. and the highest frequency components suppressed.83) It. i 0 similar. Figure 17. Thib lends t o a rcfincd nzodci for raridam processes: bandlimited whitre noise with zt r e ~ t ~ npower~ a r spectrum ~ ~ density (17.81) and finite power pnn (0) = : No 73 4 umnx (17.430 17.80) 63 White noisc is tlierefore a11 idealisation that ca. and the ideaJstlisatiomi can be justified if the white noise signal is high.78).79) and in the frequencydomain (17. it has extrenidy simple forins of power dciisity spcctrum arid autororrdation function that. Describing Random Signals From the representation of white noise in r. I)espite this. . are very iisctiil. we see that.or lowpass filtered.
85f periodic in 52. The timeaverage. However.etimc Fourier transform the spectrum (17. The power density spectrim of a discrete randorn process corresponds to the case where y = LT. As the autocorrelation sequences vS.2 discrete t i m e variables kl and k ~ .1 .6.12 A weak statioriary continuous randorn proccss Ic(t> is charactcvised by ail exponentially dccayiag ACF ) zzz ewol r .17.mdoiri process by sequenccs of a cIiser (17.[ K ] is yielded simply by sampling . Example 19. hi addition. the a u ~ o . necessarv for the tlefiiiitioit of ergodicity. Describinr Discrete Ittilldoat Signals 43 I wc will forego exaniiriirrg them in clclail. For the crosspower density spectrum of two discrete weak s~. 1%is gositive. we obtain from the ctiscret. Instead of autocorrelatitm aiid crosscorrelation fimctions. we will be using autocorrelation and crosscorrelation sequertces. i s givcii for sample r. we need to hc w a r e of tbe definitions of stat ionary and ergodicity as they apply to randorn secpienccs.. like the power density spectrum of a continuom random process with real wfues.12. ~l~~tio~ sidcrcd in the s m i P way as for continuous random processes. real arid even.k.(kT) k t IIZ The a i ~ t ~ .~ o r r and covariiknces can bc conc on.84) From here. 1 Samples arc taken at iritervals T z [ k ]= Li. it is 2n periodic. integer ~ ~ l ] drlwnd k ~ ] differences K = k. Stationaiy means in this ease that the secondorder exp ~ ~ f ( z ~ ~ orily ~ ~ [ on . .c o r r e l ~ sequence is ti~)n The power density spectrum of the discrete raiidorn process i s and i s slnown in Figure 17.~ o r ~ e l a ~ if ~ ) s ~ .a~ioiia~y faridam processes.
in accordance with (11. 11)escribiiig Rimdoin Signals Figure 17.33) Exercises xereise 17. (p).the power density spectrum of the discrete random process @zz (eJQ) is sirnply the periodic continuation of the power driisity spectrum a..432 17.12: Power density spectrum of t h r cliscrctc random proce8s q L s the mtocorrelation srquences ( p i p ( t ) .1 For which of the following ensembles might hold: .
Exercises 13 3 Ensernble 4 Ensemble 3 t Ensenihle 5: from Figwe 17. p L 2 = 0.6 The ergotiic raritlom process x ( t ) has p.2 Discuss which of L21e eiisenibks from Exercise 17.hevariance of v ( t ) = r(t)ig(t).4 Take two uiicorrrlatbdraildoin processes z l ( t )R I I J~Z ( ~ ) . E { g 2 ( t ) and v. r) TJsing the firstorrlrr expected valiies.17. Exercise 17. = 1 id nz = 1.5 Find (. Exercise 17. where p L 1= 2. + Exercise 17.~CDplayer where (lie sample fiinctioiis are prodiicrd 1y someone coritiriiially rcpeating a t m serond section from R rock C n .rZ(tf. discuss whether the proccm coiild be ergodic.c. a) Ilow large are E(r(t)} and x z ( t ) iintler the assimiptiori that the outpiat. E{. has no DC c*omponent'!  b) TJsing the firstorder expected values.(t)) = 5 and E { z . Calrxlate pg(t).1 Exercise 17.(t). 1 J ) wcak ergodic. discuss whet her the process could be stalionwy.3 Corisiclrr a xandom process ~ ( ta t) the output of i . Calculate the averages p g ( t )n$(t).7. Exercise 19. ( y L ( f ) } . E For the rmdoin process y ( t ) = xl(i) .(t) for } ~ . o.1 could be: a) weak stationary. ( t ) }= 2. Assimic that it is tlrr idcd case in which this has been happciiiing fbr RII irifiriite time aid t he CIIplayer will go on playiirg infinitely. wliere x ( t ) is a random signal with varianw = 10 aiid y(t) is any delerriiiriistic signal.
' Exercise 17. ( ~ Exercise 17. loaded die' where six always appears at times 3 N . 5 [ k ] .8 Consider the discrete random process Lthrowinga die'. a) how tjiat . where a ) r[k]is the nunibrr t h o w n a d h) z [ k ]is the square of the nurribcr thrown. value and rirri~ iiiriistic real signal d(t)'? AC power .434 17.]. A T E 22. ~ T j n ' ( f ) y ( t ) ) 1) eritails from E ( ( . Find p .7 Exercise 17. Kt EC ~ ~ i ~ ( ) ~ j with (17. and wherc ryik] is the riuniber thrown.(f) +F ( f ) ' d) p ( t ) = . the processes crgoodir? Ale Give p r ( C ) .u 5 [ k ] and n B ( z 2 [ k ] ) . DC: c ~ ~ r r ~ ~ ( ~ effeetivc.lirown at other tixms wrr ~d ~ s ~ r i equdIy. say wlrethrr y ( t ) is crgodic. Exercise 17. )  " ltc(L)* Exercise 17.1 x { t ) = e . Assume in part d) that all sample functions z t ( t )arc cvcn.e for power. But the rxinibcvs t.(t) 5E(f) For e a d i part. + Exercise 17.iice of the avwaging poiiitb? C 3 w ps7 f o (tn.9 Consider the disc raritloin process 'throwing a.10 Calculak the ACF of ~ ~ t r ~ signal i ~ t ~ Fc. Find b~~~~d am$ y//k].56). and also plJ[k. E { x 2 ( f ) )ol(f)and x L ( t for tlir deterministic sigm. Exercise 3 713 What pc~uliaritic>s does thc ACF have for the random process from Exercisc 1 7 3 Does it oiily depend oii t11e di€fert. Is thc process stationary and/or ergodic. E ( ~ )yit))') = W Y 2 (J 11 1  = n(r2(t)) + h) How can 37{x(t}ry(t)Jbe simplified if r(t)aid g ( t ) are UKlcorrelated? cf Give the conditions for uncorrelated ~ ( tancl ~ / ( 80 that ~ ) t) { ~= 0. crglJi] arid E(y"[kj). 10s).13 What are thc formu1:i.12 Lct r ( t )and y(d) be two real raiidoiii signals.li ) b) p(f) c) = 1 a(t}+ s h t y(t) = s. Describing Random Signals a g(t) = r(C) i.
Delermine u . . 4. Exercises 435 Exercise 17.17.18 Determirte the autocorrelation secpxmcc y z [ K .C. which has 0 marked on one side and I iiiarked 0 1 the other three: x[k] is the iximihr 1 lying undcincath. and (v'"). ].19 e ~ ~ ~ ~ ~ Consider the ergotlic discrete random proress 'throwing a tetrahedral die'. [ ~dzZ[&l .a). ] .7. . ~ q . ~ x 17. the autorovariance sequence [ K ] asld the power cicnsity spectriirn aTS (&) of t ~ i crandom process fiom Exexcise 1 ?.
ztioxi function of ~ ( t ) .Exercise 17&21 The sketched poxvcr drnsity spectruiri of R vnridonr I J ~ O C ~ S~ ( is ) S t givcu. Determiiie t11cl ptvwcr and aiitoc:orrc!l. .
cwistarit factor K .ioris ibed and powe~ ti~ate Tdensity sgcctrrb. and oiitpi~tof an LITIs. TW c a n use these ne\Ti tools to i i ~ ~ ~ ~ s the response of L T systems to random signals. \vC will see that the relationships bctweeii the expected valttes at 1he inpuf.S Now t h i we have ~ ~ ~ s ~ rrxridorn sigiials with corrtllation f~~K~c~.vstem take a similar Sorm to the relationships between cletermiuislir input arid ontpiit signals. To begin the chaptcr wc will deal with random signals tlitit arc' the result of xnultiplying a raiiclorn signal with a constant or by adding r~iidorii signals togethcr.ht. Rather than being interested in the precise behaviour a of the orrtpui sigrtd ct.hcory.1 crcates the new random signal . ?hen we will describe the respoiise of LTTsystems to random signals. Multiplying i i ranttom signal a(f > with a c*omplex.~i particftlarsample function of a randorri proc at the input we want to describe the system outpnt as a raiidorn process and axe looking for the expected values of' Lhe ctixtptit process dependent on the cxpcctecl d u e s of the input process. in accordance with Figure 18. Finally we will consider applicationh or the t.
7) (18. corresponding relationships are obtairred. Fbr t fie autocorrelat+ionfuiictioii of tJir output signd. Likewise the crosscorrelation Itnict ion between g ( t ) 1s ) aiid q%. ) 'P( f (18. (z) with the autocorrelation fuiictiou of thc input signal p.3) r ) K * r * ( t ) = K * p z z ( r ) .G) (18. ~7ectlt stationary raiidorn processes..{ 1). in particular.8) Figure 18. signal of a n amplifier y(t) ah thc sii~ii l tlie ideal (noiseo amplified sigxial f ( i ) or music sigrlal aud and tlie noise y(f). thc These results car1 be summttrised in a clearer form.438 18. f3.r (z) = E { y ( fi. Z~c~ause Fourier transform is linear. The addition of rariciom sigiials i s the siiriplcst and most frequently uscd model for describing noiselilw interferelice. (18.5) (18. yy.2) i s irusmcdiately obt aincd.)y"(t)} = Ejt(r = K y rr ( ).ed.c o r ~ e l ~ t i o ~ i of iht oiitpiit signial yl.z) x* ( t ) } =? E ( K x ( t 4 i  p r y ( T ) = E ( & ( l +.2 shows two ~ a i i d ~sigrials f ( f ) anti g ( t ) beiiig added to form iz new r~i random signal y(t>. .and erosscorrs4atfoa frttxbions. The properties of both randona process f i t ) arid y ( f ) are known. (18. The ideal signal f ( t ) can be a sp . We can. that f ( t ) and g(1) are joint. but the propertics of t h r suniincd proctm g ( f ) mustJbc ~ ~ ~ e ~ Fromi ~ c d .4) ) Z)X"(t)) For the (cross)power density spectra.(z) = E ( y ( f + T ) Y * ( t ) } = E ( K r ( . for example.#(r) friiictiori aiid the cro~scorrelatio~~ function plJs of thv output signal and inprit s i g i d . Corriplcx sample functions are pcrmitt.(18.z ) K * x * ( f )= l1q2pzT(?) . R ~ ~ i d o rSignals aiid LTISystems ri We now want to exprehs thc a ~ ~ t o . ~ i the rtmclctrri prowsses f ( t ) anti y(t) we only expwt enough i ~ i ~ ~ ~ r I ~0~formo ~ i 1 ati auto. represent the ordput.
13) Using (17.12): @ y y ( J W ) c.74) we can siirnrnarise both crosspower drnsitici a4 a real variable .11) (18.2 is simply Its autocorr~lxtion function (17. is therefore formed from the autoc.2. @ f f ( ? w + @ f g ( .1. Corrrbiiiing h n t l o m Signals 439 Figmr 18. Substituting (18. Ever1 wlien this model docw not rxactly repr odnre the intwfercncc in a niultistage amplifier. Tlip powcr dciisity spectrinn at the output is fouiirl by Fovrier transforming (18.2: Adtlit.56) 47w/ = E M t t 7)?4*U(t)I (18. 18. l d ) + @q.18.))(f*(t)Y*it))I if + PfO(4 + % f ( 4+ P!&) ” (18.3.12) Tlrp autocorrelation function pF?l. Its greatest advantage is the simple analysis that we axe about to cairy out for the autocorrclation fnrrctiorr arid the powrr density spectrum.1O) yields iPY?Ad = = ‘ F f M E { ( f ( t + 2) 4.10) can Bc used to work back t o the autocorrclation and crosscorrelation fuiictioris of f ( t ) mid g ( t ) .1.9) into (18.1 ~ i ~ ~ ~ ~ ~ ~ ~ ~ l Power Density Spectrum Funckion and a t i o n The snmnied signal g(1) from Figiire 18. arid then for the crosscorrelation function tmd crosspawer derisity spec*fi w n . cvith the calculation rules from Chapter 17. in many rases it still delivers a good approximation of 1 he actual relationships.iori of two random signals f ( t ) and g ( t ) Lribcd as a random process.orrebtioa ( z) functiom of t h r signals f ( t ) and y(r‘) iintl thcir crosscorrelation fimctioiis..+9(t t.f(ju) ‘ p g q ( J W ) * ) (18.2.
. Then.i)= o (18.16) and the rclations (18. j @Y. ri@vv the expression ui~~ fjd). Random Sinrials and LTISystems aiid obtain for the power density s p e ~ ~ .cviscindependent of the t r ~ ~ i i s ~signal. snrl as say long a ~ at least one has a zero Irieiin.P .18) also hold.5) simplify to PYUb3 = L7IS@) I.f ft) and . Possible corrclatiou hetwwn the ru~idoxri processes i s takm into amount by the crosscorrelation function pfg( r ) and crosspower density sp These relsi tionships become much simpler i f both raridorrr processes are iincortdatcct >~nd least oire fa a XQPO iwan.440 18.S.{(j. fyorn (l'T. the pomr density spectrim G g T j ( pis m8de 111) of the spwtra of ) the &pills .(..15) which is obviously real.t. W Q'ff 43.17) and (18. in addition.l>) t @.@) = @'ff. n t u x ~ The results so far hold for the addition of any two riklltlom processes J ( $ ) atid g ( t ) . ~ ~ kind of interference i i ~ t As this aJso usudly has x zero mean.Ol. .5) we obtain at i (@&) = 0 0 @p.J. ~ ./") + =W@. Assiirnirig that the processes arc iincorieldteti holds in niariy c a m wliere g( t ) reprcsents EI signal interfering with the useful sigrial f ( t ) . T h ~ y that when two uiicorrelated renclom processes are added. (18 17) (ju).12) and (18. the simplc relationships (18. the autoeoruelation fiirtctioiis and power j density spectra itre alqo added together to form the respective fiinctions tor the complete signal model.(P)  1 (18..q(t) wd of their ~ i ~ t crosspower density syectritm. k:xamplm arr aniplificr noise independent of the input signal CIF atitmospheric effects 011 ratlio transmission that is 1ilce.
24) holds and from (15. we can consider the CO ymncling rc+itionships tor input and out put signals of L7’1hystenis A s dcscripti forms for LTIsy5.For the cross spectrum the sarnr rule applics using t lie rcspcctiw po~ver spectra These results also become particularly simple if both random proce. then does the outpiit process also roperties? If that is the ciLq(A. B8.25) . 7’0 do t b h we fir& ( h i v e the conn the cliffcwnt averages at the input.are uiih5es and at least nnr has A mro mean 1 A c crosscorrclntion function pfcl( 7) osspoxwr drnsiiy qict’trurn Q fi..  + A t ) ) }..earid freqiiency respoiisc.. signal brings al)out the same properties m the outpiit kignal. will clioosc tlie irrlpiilse rcsyon.s do not i (’lliLwLgC2 when thc input.21) (18.acondorder expected va11at. we No assumptions are rriade aLoii1 the iriner structure o the system. for tlrr output sigiial y ( t j = A5’{x(fj). y ( t L 4At) = s { z ( t z4 At)} . f 18.12)): E{d(. signal i.iisc the svstrxri timeinvariant. shiftcd by time At (compare (1 7. we can also use the correlatiou fitriction drrtsit.v(tl).23) we obtain E M y ( f L ) ..fcms. se Now that we have clefincd tlie statistiml description of iriptxt m t l o u i p i t procwsc~ that arisc by additioii 01 multiplir+ation of each other. if the input.23) Beca..22) All results in this scclion hold cctrrcsporiciirigly lor the crosscorrelation funrtiou ancl the crosspower derisily spectrim I)etweeii g(1) and :/it). (18.I Stationaritgr an We start wit11 an LTEsystem as in Figure 18.c ( t a ) ) } = E{ f ( r ( t 1 i At). . Next it rrmst f be clarified wlietlicr the a stationary 01 ergodic iupril.y dcscriptioii that was introducwl in Chapter 17 on the output er [be condiLion of weak stationarily.(juj tlwn disapprar and we are Irft with: 7 (18. process is st:ttionary 01 an ergoclic random proccss.2..3 and considcr.’y(t2+ A t ) ) } > (18. If the inpiit process is stationary t h c ~ the .r(l2 . and oiitpitt of LTIsystems ixr detail. d t a ) ) ) = E {dV(tl + At.
26) and form the linrm mean E { y ( t ) }= pY(t)of the output signd: & ( f ) = E{g(l)) = E ( r ( t ) * h(L))= E{a(t)} * h(t)= / h ( t ) h ( t ). VI" start with thc convolution (8.s ~ i t ~ e n t ~ tionitry aml ergodic processes c m be obtairied irz tlie same way. have to be R Litatitmary process and this nieans p r ( t > and pZ. Now we know t&. Were. we w l see thel the rehtionship between the detcimincd linear expected vaiues p i t ) and p l .27). the output signal also bas the same properties. of aFnIirtsystcnr with impulse resporiw in Figurc 18. ~ ~ o ~ ~ c s ~ o ~ i t i i~ i g ~&out~ weak sta..3. * (18.4..(t) and and hjt). for an input sigzrel which is a) weak stationary or statianary. Chrnparing (18. & i t ) does iiot. but the inipulse response h ( f )is not. . ~~~~~ of an LTI To determine the Iiiiear incan at the output. ( t ) h. so the expected value fontled with ~ ( t ) p v ( t ) can be exprcsscd by pu.on~70l11lj0n 5 betwecn thc inpiit signal z ( t ) arid the outpiit sigiial y(t).rh the smie form a the c. input. Likewise. a ) joint (wcak) stationery b) joirrt. and output processes show the same joinit properties.ztionships are illustrated in Figurce 18.26) u d (18. b) weak eigodic or ergotlic.21) The input bignal I($)i s c2 raiidom variable. The different re1. (wed<) crgodic 8.(t)arc thcrpfore depexident on t)ime. The output process is therefore likewise stationmy.442 18. Rendoin Signals autl LTISystems as g ( $ { r ( t L ) } $ S { I L ' ( ~ ~ ) ) ) also reprrscuts ti theinvariant s 137.39) g ( t ) = J(C) * h ( f ) (18.
signal x ( t ) i s s t a t i o n a ~ ~ .  Response uf LTI8ysterris to I3andom Signals 4/13 Figure 18. by using conrrolution (18.F.c ~ r r e lfiiiiction at. The n i e m of a s t a t i o r i ~ r ~ mritfom signal i . signals yields . lie hysteiix cttttpttt for a station~tioii ary inpiit signd.4: LTIsystem with impulw response h(tf If the inprrt.11(0) (18.rxnsferred like the DC component of a d rministie signal.1 8.fherefore : t. and the c remaining ititegral over k ( r) can be written i t s tlie value H(O) of the Foitrier integral (9.2. The airtcworrcla tion fusictioii PyuW =E i8 Mf + T)Y*(O2 (I 8 3 l j obtained from the autocorrelation fwiction cpJT ( T) of thp impulse revporise /i(t).rd and usiiig the expected values on the rrsultiag product of the inpiit.30) with i11tegsal: Bringitig bath c ~ ~ ~ ~ v o togetheri to~form a tlor~bleintegral.1) at ih' = 0. In tht. m be Errought outside of the intrgral. first step we express the output signal y ( f ) in ( 1 8.28) m 'Flrc constant linear average /L. Thus we obtain thr simple formnla nction at the ~ u of a ~ ~ u We will calculale the ~ L ~ ~ o . the linear meal] E ( r ( t ) ) = p E is tficn constant in time and the convolution relationship simplifies to fL3 2 ~~~~t~~ = 1 w I.26) and some aim of the r ~ ~ i ~ r a n ~ ~ is ito ~ e n t s expectation opera ~ I move the integrsls and apply it to the inpiit signal.Ld?(T) d r = py . sorting t he terms l~~~ [ ~is within the intr.
444 18.33) (18.36) . R nntiorn Signals and LTISystems c r (18.
t ( ) ) * : S ( .37) (18.t o ) .t o ) =6(z) S The ACF is not (:hanged by the delay circuit.39) we obtain $q7/&(t)( r .48) as a sum.39) The mpan square a h a rucasiire for the power of the output signal is obtaincc1 by evduatiag the coiivolutiori integral at t = 0: ( 18.40) Example 18.38) artd (18. For discrete systems with irnpuise response h [k] there are similar relationWith the dvfinit.s y s t ~ ~ i the can be obtainctl with a similar procedrwe to that for the autocorrelation fiinction. (18.42) 'NIPvrosscorrcllation h r t i o i i belt w ~ c n iiiput and the output8of an L T I . signal.ion of discrete tonvoiution (12. With (18. the fumiulw for discrete systems look exactly like the expressions (18. starting with .39) for contirruous systems: ships for the autocolrelation sequences of 1he oiitput signal.18. For a stationary input.z . Signals 445 (18.1 will considcr a n iclcal delay circuit with the inipulse response k ( t ) = s(t . (18. Resnovise of LTISvstems to Random.2.38) (18.41) (18.35).
46) By substituting 7 = /A we finally rcach the desired expressrioii ibr the c'ross1 correla\.44) then jntertahnnging the sequence of integration and expected va1iir.58) yields (18.r) * p.46) we obtain (18.38) The cwrijngate syininetry of the antotorrrlation function ( 17.(T) = / h" ( .47) U Tlz~ fornirila obtained for thc crosscorrclntion function pg.49) Both possible crosscorrelation functions between the input mid outpit of tm EHsystem arc casily obtained by convolution of the auto.o rcmcmber.z : ) (18. Froin (17.correltition fiirictiori at. response: (PTI. is easier t.51) (18. z (18. (z) .54) in cwnbimtion with (18.52) I I . LL (18.( r ) betmmm the inn put and.hc synimetry relatioriships for correlation fiinctions from Chapter 17.) dv = 11" (. Hurzcloln Sigiials and LTISystems Llze corivoliition integral is irisertcd (18.446 18.v ) pr5( .47) we fi~st i a k ~ iise of t. outpiit.s yields 11 the expectcd value we again recogniw the autocorrelation function of t hc input 1 signal ~ ( t ) : 4Sd4 =/ fwPrJ(T+P:dP.50) (18.2. the input with the imprtlse response h( r) m c l h* (.4.ion function of the input and output of nn LTIsystem a i d i t s inipulsc.a. To derive it from (18. .
~ ~ o ra at arid its freqricney response. We can obtaixi it more quickly from (18.yclornain description. arid with (18. ( 3 W j I f * ( % J ) = (@ ' (18. (18.56) r1 ensity ~~~~t~~~~ arid 15Je I i w e iiow dwived the relal iondiips hctween the correlation fiirirtions at the input m c l omtptit. = of I ~ ) } (.571. They turned oiit to be verg similar to c o ~ i ~ ~ o l iof tis~ o ~P r ~ i ~input sigriali and impulse reqmisc.70).[ (18.spectrum with the frequency response of the LTIsystem.p) py5( ) } .52).(J4 3 1 . with (17.70) yields = 3{ r t @. Foiirier r> transformation a. 8 of randorir signals.f=(CTTIZf(J).~ i c the ccmvolutiori theorem (9.52) filld (i?Y&(d =ii(Zo)*it7. 'This sii?tple insight is E used in tiiany eiigineeririg systems to measure signal delay. sponsc iz[k]holds p&i] = h*[K] * c p I 1I.54) PtlJ. As we have nlreedy rnPt the powcsr density spectrum a a freyuciic. j d ) 3 { ~ r z ( h e ACF (17.. hcmever.[ I = h [ 4 * PTLLl4. for PxarripIc.55j ( L8.53) (18. T h e ~ r ~ ~ s ~ . using Forrriw transformation.57) ) The crosspomw dcusitp spectrum @ v s (jd) is obtained froin I he powcr dctisity spectrum Q F T ( J W ) of the input signal bv multiplying it with the frequency ieqonhe H ( IW) ol: thc L"1system. The crosspower density sprctrnrn @ L y ( . WP will start with the crosscorrelr~~irra function pYx( as in (18.~ ~ ~function t between tlii.of LTIsystems am1 &he iiqmlsr rcsponsr.69) and tlre CCF' (17.1. ? * ~ 1 ~ ~ ~ J W (18.74) as @.. input txnd output o a delay circuit iz rrel~ ioii f simply a shifted version o the autocorrelation ftmction.18. ~ et i ~ ~ ~ ~ c An even sirriplci clcscriptiaii can be obtained in the frecluencyrlomi*inby rriultiplving the inptLl. Response of LTISystems to Random Signals 447 We again consider thr Ideal delay circuit from Example 18.p(~. ~ coiilti also be derived in the same ~ ) way from (18. i t is reattonable to i ~ s i m i e that there are some similar rclationships between ~ c ~ ( ~ s sdensity sp ~ ~ c ~the input and output of a systeni .58) .2. r ( .  Rrcording to the c o ~ t ~ n u o a ~ e for a discrete LTfsystem with irnyitlse r ~ cu ~ .51}.(JL"! =~ ~ ~ ~ . radar or sonar.
448 18.lad output of BI~ LTlsystuni steps as in the dcrivation of pl/J r) in Swtioii 18.2.~icfomn Signals itnd LTISystems Illere wr I ~ t ~ used the fact that.4 wotild then hctvc l d to .4 (Correlation of 1l)eterministic Sign&)..73).60) The significmt rclationships for correlatioii fiinctions aiid power density can now be summarised in a clearer foriii: Figure 18. ve a).38).. after thc results of Chapter 9.u(~w) it simplc rdationship can be fourzd by Fourier t r a r i ~ ~ o r r ~ ~ j ~ i Thr Fourier t r ~ ~ ~ s € o r thf~ (t8. Ka. 1 x 7 expressed by the magnitude squared frcquency req~onse (jd): N (18.2.9. P'oi the power density spectrum of tlic output signal ?f?T. ~ of r r ACF q j b h ( r) can.5: CCF of M signal with i h e input . tlic power density s p ~ c t r n m . ( j w ) is real (17.
Thc change of tlie iiitlex with time convenitwtly prochces a tiiffereritiator witli tlic tiaiisfer fun(tion The lcsiilting dmivative y ( t ) = + ( t )has the power tlwsity spectriini .2.18.riiis to Randum Signals 449 wfiert’ U is very siuell. Response of LTISyst.
3. Figrxre 18. this is not ncccssarily a problem. instettd of thc coi~volutioirintegral as in (18. ( g ~fiorn (18.3. we calculate the rpadratic mean of the output signal y ( t ) = h ( t ) * ~ ( 1 ) .73) (18.rum) witli i~ thought txperirnerit . and the bandpass thtvfore lcts through o d y a vr1y ni1rrOW frequpncj bztnrl. We stnrt with a iandoai signal ~ ( t ) whicli has power density spwtiurn @ 2 z ( ~ u ) .72) 14% hrme already encoimtered bandpus filt ers of this kind when we loolrrtl at sanipling complex handpass signals in Chapter 11.1. arid thry produce complex output signals from reill input signals.450 18. I3ec. Rsidoui Signals and LTISy&cms retation of the We can cxplaiii the irrterprekation of thc power demily spectrum (which should actually be called thc autopower deiisity spect.6: Baridpass wXiich has been drifted to the right. It will Le fill c ~ r d through an ideal bandpass fill cr with frequency rcsporrse (18.75) .64). and i s therefoorc complex To delcririine the power of the oritput signal. We use !he inverse I'ourier trmsforrri of @ g y . IIowevt.40): j) (18.iirw tbcy do not ha17e conjugate syninicfry they have cornplcx iinpukc rcsprsnses. wc will imagine that Aw is very small.
LCesponse of LTISyst eins to Random Signals 45 1 which we gave in Chapter 17. O w c e s p i d i n g nicwurement setup ic. the amplitiidc and phase of 1 he oirtput signal.5.2 without proof.66). Tlir v r 7 1 d white rioist. in prixiciple.put respoiise brit. The corrclation between tliehe bandpass componeiits i. A l)ett. ( j u ) between a signal r(t)mid a referenct~ signal r ( f )citl1 bc carried out similarly. (18.4): p r c (r ) = 6().xii thc Iranclwidthof tlie systrrri undcr investigation. it will appear in the cross sp riini.rr. X i thc reference signal r ( f ) i filtprccl with a corresponding hmdpas? filter. Q.5. 0 4e.. This approximation is always justified if the baiiclwitlth of the noise so~irc(‘s miich greater th. riot quite so descriptive. in practice. because otherwise it would ialtdfy the xrrewuremciit.(. a impulse causc many pioblems. The i urikriowir tramter behaviour is obtained from the crosscoi1elittioi~bctwecli the input and out put. Tlieoretietltlly it is possibic to excite the systriri with a delta impulse and nreiisure the oui..62): . ( J W ~ turns out ) to bc the C Y O W spectrrun aV7W O ) of (J equcncy coinporient g(t) = . a. morc modern niethotl 11 R wideband noise sigrial and strctistical analysis of thP output signal.2.. Th tw arc various ways of uneitsuiirig tlicm.hangc1.18.. recoided iii the crorspower tkwity spectriini. the high ariiplitudes rcqiired for an approxiniatioxi of tlie d ~ l t .. Again having in uriiid tlie analysis of z ( t ) into inmy IiarIotv band pass signs ccoidiiig to (18. ion of t h e crosspower density spectrum Q.I (t) * h ( t ) with frequenry u ) ~ ) . For white iioise witli noise p w e r N o = 1 (see ChaptAw 17.76) For the crosscorrclatiorr function pgJ( 2) 1)etween the output thc crosspower density h l t w t r r i r i i a ( J C V ) follou~s . . signals...)d) = 1 . * i d )= 0 ( The transfer beliaviour of an Ul’lsystem is corryht ely (lcwriherl by its impulse response or transfcr function. c. shown rid nieasiireinent sigrial at the input is cl in Figure 1&7. r . Beyoncl that the frequerrcy of t h e siriiisoidd signal r i m only be cliaiigcrl slowly ciiough that no mtdesired traiisients appear. If there i s a fixed il1rlplitudP ailcl phase relatioiisliip bctwcwl x ( t ) aiid r ( t ) . t hr oiitpnt signal may not coritain any int erftwncc. from ( 18. Clearly oiily cornpoiwnts of tliti ham(’ t h e crosspower spectnuxi do frequency c~xitribiit to Ilie crosscorrelat ion betweexi two e ioiiarv signals. A prcrequisite i s that apart frnm i t s response to the sinusoidal input signal. Another way is to excite the s y with a siiie wave with varying frequcacy aiid xileastire the frqucmcy r c s p o ~ i swit h ~. If the frequency co~npoiicnt s do not correlate. evtln thougli the rcsixlt i.
a n iiiterftwnce signal (noise) wil h n rnndorn cliarac ter.452 IS.2 to good LISP in in Figure 18..par ( r) ant1 crosspower densit) spcctrurri (id)iiizmediately give the impulsr pmrsc and transfer func'tion of an unteni.7: Analysing an LT1svskra with whitc noise ( 18. The filter H ( j w ) will be dtwloprtl so h i t i t s cru(pul sigiittl g ( / ) corn to the oiiginul signal <LS possible Tbc task of thc filter i. Random Signals and L'I'ISystems li'igute 18. then we will he aiid i>blcto p t tllr thcory \ye ~levelo1)ecl Swtion 18. The potvcr rlensity spectrrtm @zr (?U) the observecl sigiral can oi rimnt.ruction of zt signal that his bern ~ O Yuptetl by "itTioLls influencer. The inflticnte tvlikh corrupts the essiblr. for cxtmplt~.TIwill systeirr and tidt1iI ionally. \T7e will assnrtic t h t we oiily krrow the crosspower clensiiy sp b e l w r n the original signal s ( t ) arid thc obseuvcd signal x ( t ) . rerndrrs mrc~lranged. If TW are tlcvding with weak statiorrary raiitlom p~~ocesscs LTIsystems.x long as . uud is this also k i ~ o m ~ i .tl hy tnrasriremcnt at any tiuw. from rrieasiiierirent with R known origiirtil signa 1. 'I'hc c*ros+po\wrclciisity spectrum is a rnatbematical lormixl~tion (lie infhirnc'e lor t Ire coiiupting inflnrnw oii ih: of original signal. only the sig originwl iligrral bcforc it is obsr~~vcd be described with the p r o p r i ies of aii 1.r(i) itrid ~ i j tdo iiot correlate (Excirrwc 18. i ) ter ' Y h field of bigad est iinat iori tcmcerris reconst. to eliniirmte . No details about the structirre of the 111 Isystcm arid the iioisc sigrial arc known.78} The crowcorrclntion firriciiorr . The origirial signal b i t ) is not directly The problerrl i\dry (1) call 1w obwrvcrl.17). TjLle 01igiriztl signal is corr I ed IS nillch iks pobslble using knowldgr of the statistirnl properties of the coriuptirig irifiuencc. tlwretorc. If the sigiial y(t) i b lapped by ZL rioisc signal n ( f ) tlie eesnlt .8.
so the received sigriat ~ ( t110 longer corresponds ) to the t r a n s ~ ~ isigiid ~ ( t f~ i t ~ s~ . wc’ 10 . Thc signal z! read at a. and t he interference Irc~e also be iiroddcd as noise. Rstimation of this ltirxtl is irwc~ssaryin niariy different applications as thebe exan1ples shorn. Signal Estimation Using the Tn7ic11cr Filter I obse~ation 453 F f Figure 18.. Signal transmission If a signal s(t>is to br transmitted to arrolher toc~ation.(jw) and CP. ~ a In this case s(t) is a physical quantity wbosc t i i ~ ~ b c~ i ~ vis o u~ ~ ~ ~ ~ r i r nieasurecl. tlie measiiiing clcvice gives an ina iiratc observed version ~ ( f of the original signal s(t). Inter tiine () cont ains influeiires from the rewr (ling nrid reading eqiiipnicnt . In order~to stow a 0 signal s ( f ) . it will be affwtetl by the xioiiideal properties of the Iriuisniission chain and may aIm be sixbject to i n t ~ ~ f ~ r e i ~ ~ ~ e . filter.rotlixce a yneasut‘c tor the differmcc between a(f) and ~ ( t )called thc crror . In the following sections we will develop it mcthod that provides a s u i t d k traiider function IIjp) from tlie linown yuantit irs @. ) Addil iondly.18. As each rneasurenieiii aflects the process being measured. ( J W ) in a systernatir way. errorLj in ineasurenient can be niodelled usiiig noise signals. The filtsr is called aii optmu1 or Vl/zCnC?.~it must be~ h n g e d so that it romplies c ~ ~ with tlie rcquirenieiits of the target metliim. ‘ aehievr t h i x .3. power or the estimation e ~ r o r + s(0 noise I I filter ~ ( j a ) ) Y(X .. (‘ail ~ ~ 1 To derive the optimal transfer furivtion N ( j w ) we first of a 1 need to find a rnaihe 1 rnatical approach which shonld preciselyf o mulate the requircmunt thst the output ~ sigiinl g ( t ) is ‘ah similar a8 possibk’ to the original signal s ( t ) .8: Estimation of the o ~ i g i mxigrrd s ( t ) l the inflwrrce on the original signal. xiill int.
82) H that leads to the smallest error p o w e ~ i. . To determine the IiansfecJ.2 andoin Signals and LTISysterns Alt hough we do not kriow t.5. (.2 that the Inerin sqime E. 1 2 Bnow from Chapter 17.we obtairi @PC I = @yt. lising (18. tlepeiitlent on lf .ju) the output hignal and @ .79).9). ~21ld t o make it as small as pos4blc.82) by ~ l i ~ ( ~ ~ with ~respect to H . = EH"  l€ a.87) .5 (18. in (18.?U> expre try is thc power deiihity spectra @ui.64) arid thc general ~ ~ l ~ t i o (t8..81) for the addition of random variables. .13). ~ shouldi be iiotcd hcre c ~ ~ ~ ~ It ~ t ~ ~ ~ that tlie variabln.~w) tlie rewiistnic01 titm e1'ror c(t)..  N"i  @.lx onginal signal s ( I 1 oi the I econstrnt~t error e ( t ). t +'t> (lt3.. ion we can exprebs its power with kiiotva statist3icalqumtities.% culaled by iiitcglating the power tleiisity spectrum arcJ W ) (compare Figurc 17.82) dcpendcnt ori thc iaegnitude /Hi and yliase t$ of the freqrxexicy revpnnse T h e estimation filter N = jH1 ejcb (18. rxprcsscd by the correliit ion with 5 . Xlere the argiiixierit (jd)is fcft o i x t to siniplifjr Ihe notation.454 18. Q.67) with y can ire. obtaiiicd from (18.$9. ( ~ L J ) of the original of signal.83) (18.82) :ml H are both romylc 'Tlwreforr ive write (18. t.(.r finrtion H of the estimation filler.@ys . Qt. ( JTc will Lbercfore consider the p o ~ v c c density spectrum (lzLC(. the ~ ~ ~ r r e l a t ~ o ~ i r i s ~ and (18. we need 811 expiession for tlie power density spectrum Gee..{ / t ( t ) / ' ) cali d s o be cztl . From (18..66) ~ ~ ~ ) ~ with 'P = s. with exprcssion (18.
Tlre result car1 be iI1nstratd h y irsiiig l R coriinmi niodc>ifor signal inl cr fkencc: f i c m deterniiiiistic arid rxritlorn sources.. and (18. This rstimiltion fi1te.8.1. a.83). The interfcrenc in this case mor c accurately motiel led by Urlspstcm with frrqiicncy respouse C(JUJ) 811adtlitive noise s(urce n (t).3. Tlie freqnenry response G ( j w ) caii.tlrc complex frcqueiicy response of tIw cstiniation filter caii be o b tairred from (18.64) arid ( I S. a transrnissiori cable.9 1) i s called a Wieiier filter. First. The estimation tilter given by (18.GG). Anyway.ecorries zero when thc ptiasc is T h e phase response of the rstimation filter is tlwrefore tltoserr to br cqrral to the negat.1 Linear ~ i § ~and Additivei Noise~ ~ ~ ~ o T h e derivation of tlie tVie11ex filter as in (18. we fkid With (18. The mioise signal may not corrclatc with tlir otigiml sigiid s ( t ) . acmrding to Noxbert Wirner (18941!)64).91) is gwicra1 becausrt we requircd no knrowldge of the k i d of ~ i g n ainter feientv.rr.91). because from (f8. drc description of raiidoiri proce by mean values xud pomw sp construction of ttn optinial estixnntion filter..r corrects tkic observed signal ~ ( t ) thal the deviation surli of the result y ( t ) from the origkial signal s ( t ) exhibits the smallest c*rro~ power possible.sscorrclatiori fiinction The magriittitte 11 11 i s iiow i h complrtcly defined. for arid cxample. stand for the fretpericy rcspmse of a n amplifier. We dcrix7e t h t ~ frequency resporise for this setiiy of the Wimer filter from the gerteraI casc ixz (18. Figure 18. All evtcrnal iriterferewt~ sources are siirriixinrized by thr additive noise vigual n ( f ) .90). using (18.G (1832) .86).The optimal phase Q can be derived from wliere tlie imsginziry part 1. we obtain = @.89).is in Figure 18. 28.9 sEiows tlic same setup . pioncer of est iniation thewiy.89) arid (18. We derived this result without knowing the origiiisl sigiial s ( t ) .im> phase response of the I. or a radio cIiannel.
TL7L(~w) the noise signd. frequency response G(2w) makes i t possible in this case to find froin the the crosspower drnsity spectruxn Q h r (JLJ).456 18.91) . Raridoni Signals arid LTISystems Ecx~o~vinp. the power density spectrmn a.. of original signal ri (18.nd Q. ( j w ) of thc original signal a.
18. and the filter allo~i's > ) the signal and the lowpower noise to pass t lirough unhiirdered: H o d ) M I . @. (jco) I 0 Figurs.10: Filter for removing adclitivc iioise froiii a signal . 1f both pon7er density spectra are ot thc same order of niagriitude. Sigrial Estimation llsiiig tlie Wieiier Filter 457 I I Fignre 18. origirial sigrrnl.. j ) of the Wierier filter. @ . mid the lilter halms both tlie noisc and the siiialler sigrial conrpoiiexit: N ( p ) M 0 . 7 (thcu ) << .3. s For ( a . aiid frqncncp re5pon5c I I ( j . @ . l r L ( ~ ~ the signal power domiriates. 18.11: Power dcnsitJ spectrum Q 5 < ( 7 w ) the. of noise. the freqiicrrty resporise of the filter takes a value between 0 arid 1. L 7 1 ( / ~ ) ot the . ~ noise dominates.11 shows ail cxmipltl of an original signal with R stIoiig1y frccpiicydepericlcnt power density spec imr arid a noise soIirce that produces whitr noise. Tlic fillrr lc%sthe frequency components pass wlirtP the original signal outcveiglib 1he noise.>(id) @ 7 7 . and stops the heqiwiricJs wyhcre thc original signal makes no contribution. ( p ) . . * For Figure 113.
this will be sucwsdul. . xcises Exercise 18.y response of the Jfiener filter 1Leu takes the form T r (18.antforn Signals and LTlSvsterns Linear Distortion Without Noise The other spec.1).3.__   I .itability of H ( j w ) (Chapter 17. it was riot cpiite complete.hut bcrituse the noise was disregarded.53) aid (17.458 18. ‘I’hc following syst rni has coiiiplex constants A and B anti a stationary corizplex input sigiial ~ ( twith ACF pAL.3.2. In addition. anti is ~’vtw intlrprnderit o f the power tlcnsity spect inai Q. R.G( averago / A .2 ‘decorrvolution’.98) which nteans that it is trying t o reziovr the effects of the distorting systmi C(JW).5.1. C a ) using the definitions of tlie CCF and AC‘F’. ~ The origiiitil iignal s ( t ) i s then only coirupted by the IlI’lsystem with frequency response C ( J W ) . 7 z 7 z ( ~=) 0.1 and 18.2.3. (17.h e frcqwnt.2 xercise 18.iaI case that we can derive f*oiii thc freqiienry reipoiise (18. we have to bear in rriiiitl the .96) of the Witner filter is ideritilirtl hy rcrnoving thc noise so~irce: Q . ) z) aid I __ __ . s s ( ~ w ) the original signal sjt).3.1 A complex constant C will be added to a slationary rsntlorn signal s ( t ) with average p arid ACE’ p1I ( r). As long as IG(jw)I # 0.3 18. We have already distussed of l h i v c i w in Chapter 8.56) b) with Stctions 18.
. (1) and e ) with Section 18.61)( 18.3 For the system frttin Exclrcise 18. vV7. B E IF?. verify rwults a). Exercises 459 Exercise 18. A. The raiiclorii processes u ( t )and v ( t ) i i i c stationary.18. aiid . r .... p .find the 7) and iii c) Ihc CCF b e t w e n d) the CCF betwern U U and aiid U’ and 71 e ) tlie CCF hctweeii 111 Use only tlie tlefiriitioii of the CCF for c o m p l ~ x signals (17.. r ( t ) and y ( t ) . Finally.64)! Justify your answer.( I).!(z ) ... and .2.1._____I F~om giveii coxrelation properties Y.( arid y.53) and the vxpressiorz (17. Z ( t ) sZZ(r). . Find q 2 S ( t ) . Exercise 18. find gszl(z’). a) Usc the definition of the CCF (17. and complex constants A aiid U. 1 L a) Lhr ACF of w ( t ) b) the CCF hrtwceri U _ ..4.54). Exercise 18.r(t) has a w r o mean arid y ( t ) has inean p.~.53). 11) Can thc problein also be solved with (18.5 The linear combinatiori z(l) = A s ( i ) iBy(t) is formed from two independent random processes .4 Consider the following system with 2 iiipiits a d 2 output. have Lero incan and arc cornplex..
8 11 1 has t lie transfer fuixAim a) Deterniinr t hc impulse rcqxmse h ( t ) b) Find the filler ACF phh( c ) c ) At the system input a signal is applied with power clmslty @ I ( p ~ ) = No rn6(w). r ( t )is station.iry arid Fiiid ttas a z ~ r o . timeinvaxiarit system is rlescrihed by its t iansfei fiinction . a i.r(t) arid y(t).6 A linear. of thc output sigiid 7) !At). an input sigrral r ( t ) aid output signal Ytl).2s + 2 N(sj = ~ (b 2)(s2 + 2s 2) + + Its impiilsc response is h ( t ) a) G a h l a t c +lie filter ACF phh() and skt T b) White noise with power density No . = a ) the power tlensity qpectruir. . Givr the autororrPlation function pyT..1 is applied to the system input. .( c) :ml the mean + d) For t.si(l).~ ~ i e qir s ( r ) S(z) and k ( f ).he oixtpiit signal y ( l f ...( and the power E'. find the incan pY.s2 .the mtocorrelation ftmctioi? q u 4 ( r )the out pnt signal power PV aird the I W L T ~ P density spectiuin @. h ( t ) .460 18. Rautlont Signals and LTISystPnis Exercise 18. F'inct the ACF of the input signal p..7 A system has the inrpitlse rrspoiist.(lw). ?bh(c) 4P Y Y W (F?dT) f ) Poww and wriaitce of . Exercise 18..tts arid p?.1of x(t) b) . Exercise 18.
14 p is often wed to gcncrate a signal with the spcictral properties signal.41) from the ACF of complex random series y r z [h] = E ( . . Exercises 461 xereise 18.54) from the CCF of complex random . I power deiisity No. Thc autocorrelation of the output Iy)$CnlZI.( = z) riiiine a possible tiansfer function H ( s ) for the system.( z) and p c) Give the corresponding crosscorrelation fuiiction j ~ .1 . L [ ~I C ] X *[ K ] ) . the mean and the variance of the output variable y ( f ) .3 2 Derive (18. Exercise 18.9 A systeiri with the transfcr fuiiction H ( s )= s1 s2 I 3 s 3 . white noise sig~ial ( t ) with 1 + y i d is measured ptl.iiptions. Determiiie the autocorrelation function. b) Can the transfer function be imambiguonsly determined? Exercise 18.10 A system with transfer function H ( s ) is driven by . >\liwe p.. ) a> E x p ~ s s autocmrelation functions yyiyt and + y z ? .4.. Exercise 18. Exercise 18.cries p s y I ~ ]E { s [ k t = K]J’I [k]). ik.They both h a w a stocliastic inpnt signal z ( f ) with autocorrelation function y p z ( Two stochatic oulput signals yl(t) and y ~ ( t are produced. I .. where (P > 0. z).13 Derive (18. ?. + Exercise 18. b) Determine the crosscorrelation frtrictions yyl..2 is excited by white noise with power density NQ. ~ ( r~) .&[k] . L ( z) of the outpul the (71 sigiials with the given signal and system desc.111 Two causal LTTsystems are described with their inipulse responses 111 ( t ) and ha ( t ) .18.
15 A transmission clianiiel is cliaractcrised b y the following system with real constaiits U and b: I   I The interference n ( t ) does not correlat.md ( in Note: treat the slietches in the barne way as Bode diagrarils. Exercise 18.g. Ir I G ( J L J ) ( ~ for coiiiparison. Exercise 18. n( t ) .7L(jwj No arid @%Jju) 1. as shown ill tlie illustrnt. ( ~ w ) ..h of H ( j w ) should be denoted by Z(t). a) Which statistical signal properties must be known to solve the problem? Give H ( j w ) dependent on these variahlcs. sketch t Solve 1)) for N o = (3999.errriine the tmrisfer function H ( j w ) of a system that reconstructs R sigrial :c(t) from ~ ( t ) minimal mean square error. o plot the cuIves x using a cnniput er. also plot Qn7? p ) t . . Raiidoiri Signals a i d LTJSystems a ) Calculate the freqiienry rcsponse I I ( t J " ) the system with outpnt v [ k ]arid of 011t p11t s [k] b) C'alculate Qyg(eJn) and sketcah it (e. Both signals are white noise with pcmw densities @ll. Sketch @ P s a ( ~ d ?)v.162 18. I N ( p ) I and < w < 104.ioii. Is the solution o p t k d with respect to thc task set? b) Find H ( p ) as well as cDep('w) for the\ transmission chsnnel shopin above ~ for n = I arid b = 100 Set Nu = 0. v ( p j . = = Det. with MATLAB). @ Z r ( j w ) logarithinic form for 10 in L) (1 ol @ 3 5 . T h e reconstructed signal at.e with x ( t ) . the output wit.16 The trRrisniission of n sigrial s ( f ) is subject to interference from noise n ( t ) and distortion from G(sj .
Detcrinine the Vl'ierier filter H ( 3 .(jw) .n(t)is rneasured with the arraurgenient frorn Figure 18. cli. Excrciseu 463 'The signals s ( t ) and t1 (t) are .18.7. .4.. clepencling on @ t 7 T t ( p ) arid G ' ( j w j . j w ) i. correhte. xercise 18. m Show that @?.~iidcr the iiifluence of interferencc .27 T h e frequency rcsporise H ( . a) gwcrally.tincorrc4tttd. ~ 'that will try to xecorrstruct the oiiginnl iigiial ) s ( f ) from ~ ( t ) .:r(~u) j w ) holds when cpzx(r)= 6 r ) and when n ( t ) and = H( ( z ( t )do not.
This Page Intentionally Left Blank .
i.. and i s a digit. digit a1 linear.2 ~ ~ ~ ~ 0 ~ 1 II a hard disk is considcicd a “l)l.c. If. rontiniioust irrie. digital tl) cliscrt. a digital signal i s satcd on it as a srries of oneh and zeros.e. rausal .. riot digital c) contimtoiasarti~)litilctt.teaxnplitiidc. ~ o ~ u t 1. riot>digit d e) coritinuousaniylituclr. mrrnorylcss. neither analog nor digital.ick box”. analog. ab it i s cwritiiiimiis in time arid amplitude : analog (althoixgh s o d y changes it15value at cwtain poirits in tirnr. it 2 is defined at every point) amplitude. i i o ~ v w the .. The write voltage has discrt tude but is rontiriuoustime.inner workings arc considcrtd. tliscrrtetime. i.e. analog. with rneniory. it mist be more precisely de~~ fined: The hitstream to br written is digital. tiineinvariant. contimioiistiriie. i.. discrctctinw f ) eoiitirii~ousamplitndc. digitd bj continuons.~ ~1. continuoustime ~ 0 ~ 1. so it is rieitlirr :in arlidog nor a digital si magnt. not.1 l ~ ~ ~ o ~ a ) discrcitcarni.3 ~ io a) t’l 52 : analog. discr tinre. The read is convertadby rewvering the clock signal and a hard decision into a series ol: ones and zeros. caiisal System 2: iionliiiear..inuotmtirne signals.al signal.tic field strength in the disk a i d the read voltage are coiitiiii~ousam~)lit~~cle and tont. timevariant.litiidl.tni~~litmctr.
c(t~)}=z(t?'~(f))g ( t . timeinvariant.c(t t tT  r )clt' 7.i nvariant as S{a(t . cansal as the outpiit i s equal t o the input signal in the future' with memory.4 Aripencbix A. e ) see not. as the 1cs1)onse t o the iiipiit.I)} . I tT . causal c ) liiiex timeinvariant.466 olution 1.z)) . timeirivariaiit . tiimiiivariarit. the ontput signal shifted by z: S { s ( f . wit 11 meiixxy 11) linear timevariant asS(. signal points in tirntx c) lincar f ) 1' inear tin ie.z .riot causal. Solutions to tlic Excrcises a) linear. with memory rannot br rcalised =.= + J .=f'. as the response to an input signal shifted by z is the same a. d) 1' inear tirn~iiivariaiit. causal.T ) = y(t..c(trT(tz)) . 1 It(ll)d71= ?/it I) z callsal with mcwiory as all past valitcs from t = 0 arc' saved g) linear.r(t.r) = causal.z ) f causal.. timt~variimt. with mernory i) liiicar. also cittisal h) nonlinear. inemoryless. nienioryles4 and thewfort.T  r . as tlic output rlom riot depciicl on future input signals with I Y I C ~ O T as the delay requires t h t the signal i s s i i v d ~.
however.r) . b~ ritticr one of the two.Appendix A. barne h i e . olutiori 2.h.o t. know from 3..6 vlic. arid it is not possible to say w1iic. the response to x2: .oiild. that tlir system is not linear and tiiwinvnriaiit at tlu. It c. QL 2=0 Substitutiiig vmiahles 1' = 2  1 + .he Excrciscs 467 ~olutiQi~ 1.1 a) Show that. 1)) Tf 9 is the spstwn response to r1 and 1 s/2 dt d ( t .1  dt 3 dr' = df'.= rlt' the sulxtitution leads to thc above equation. Solritiolts t. ie.
tlform is found by interchanging the left.a~ionic. ~olutio~ 2. It does riot liave to fulfil1 any c~onditionras dif€erentiators with tlic saint input signals have the same outpiit sigii.1 ailcl h. ho L 2 0. BJ setting the cocflicients CIO = 0. tvhcrrlinimllni nuniber of energy s t m w (integrators).4 a) Block diagram b) A c. 112 = 0. a 2 = 3.t Is. 2. Solutions l o tlie kkcrcises Solution 2. CI i = I. Solution 2.5 a) 'I'hc stitteqxwe representation is given by z 100 2 J 100 0 0 = [ 0 100 ] 2. a n d right sides arid using a cosiiiiioii differrntiator. = 1.5.468 Appendix A. .2 form I T is canonical as it 11.3 Solution 2. we fiad the block diagrams slioxn in Figs.1 H l l d 2 3. ( t l = 0..
W e clioosc: z 1 at the outpul of the left iritegrator. a signal flow graph ran always he clerivrd from thc statespace represelltation. f ~= r 0.100 into Fig.1. at ail irltegritt(j1 output. not observable from arty of the outputs as every row of C lias a zero. c) Insert {lie coefficieiits c OJ .6 States at the outpn(s of the integrators. u2 = 107: u.3. Solution 2. Parallel for1rr U # 0 arid A # 0 Not cornplelely observable.ADpenctix A. a1 = 7. x2 at tlic output of the right integratoi. with a modal form of A that is not in geneial mambignous. d) Both arc caiionical. Solutions t o the Exrrciecs 469 If' every state is piit. With T  2  we obtain: .7' e= [ a 8 i d=2 b) 'rbc transformation will br carried out. 2. arid an adder is connected to evc~y inttxgrator input.3 = 200 ancl Solution 2.
(1) 110.3 a) Corisistmt norrrialisation for t hc components: 1 This yields Lhe iiorinalised component v:ihic+x R .)le as all elmwits of b are nonzero. i s conLrollaI.lernents of i arc iionzero : 0 1 0 0 1 0 . Soltitions to tlic Excrcisc.:. L = 1. Thc systcm does not ckimge its inputoutput hrliavior with thr tramformalion from Eq. and right iiiiiltiplication with T as vertical mirroring. aiitl it is ohscrvable a.2. b) no.47 .50.1 Solution 3. b) The proof i s easy if one recognises that left niultiplication with 1Cl can be intrrprf>tcdits horizontally mirroring the matrix elrinc'nts.1. C = . The cigenvalues of the system air 2 Tj (equal t. as H ( 5 = j 2 ) = 0 Solution 3. . e==[ 138j 13tsj 1 d=2.2 6' .2 a ) yes. Solution 3. 0 0 I .o the diagonal elements of d). c) no. c) It... e) yes. 2.470 Apprritlix A. all c..
.4 Where 1 < 0 I y(t) = 0 = X .Appmtlix A. z ( f ) Since iii the swond rase there is cigmhmction of the irit cgrator. 1x0 X to satisfy the condition. Solutions to the Exercises 471  Denormalising with t in s aid 7~ irk V:u ~ ( t ) .V e . ~ ( t is not a n ) .i f (:OS(t) = 4 S I 4 Solution 3.
C ' = a . C = 1.x(t)\ 5 A l e r r for t 2 T . C c) yes. = 0. C = 5. e. T =0 a) h f = I . 2' =. witli M . if D > c'. if 1L1.e. T  =0 =0 b) yes. e.g. e. r' ~ol~~tior~ 4.0 e ) no f ) yes. i.472 Appendix A.g. C = 6 .3 First determine M . The bilaleral Laplace traiisform exists if the regioii ol' corierge1m is not rmpty. with AI = = 5. with M = I.I . T .r(/)l5 Mc'" for I 2 T anti l. e. G. w i t h ill = 7 . izritli M d) yes. C and 7' can found such that: I.0 1. Solutions to the Exercisrs ~ o ~ u t4. D = O .r(t)l 5 Me"' for t 5 T. T = O =+ R O C = { ) .g. T I. e.2 ~ io Rightsitled functions have exponential mxkr.g.g. a) yes. D a r i d T 50 that /.
4 ~ ~ n t Eq.Appendix A. 4. + 3)(s 16s ) ( s11 1) +a t  (s + 2)(s + 3 ) 5 s i 11 ROC : 1?~{5} > Solution 4.4.’dt ~ + x Sol~ti~n 4.r(t)e’t tlt L ( A f ( t )IB g ( t ) }= / 31‘ [ A f ( l ) L ? g ( t ) ] ~ .1: C { x ( t ) } = X ( S ) = x 7 .1: G{:r(t)}= X ( 5 ) = . Solutions to the Exercises 173 ~ ~ l u 4.6 EL].5 a) F ( s ) = b) G(s) = c) F(s) 2s 3 __ (s 2 ) ( s + 1) + + ROC‘ : R e ( s } > 1 ROC : X P { S ) > 1 ? (s 3s + 1 + 3 ) ( s + 1) 2 5s2 + + G(s) = (.
8 1. first scale transform pair 1 with n 1 iii timc: ~(t) o $r . ~ ( =) ( t ) t ~ 0e 1 X(S)= . X(S) = ___ . Solutions 10 thc Exercises olution 4. see Eg. 4. s +n R { } > lZc{a} es = HI. see Eq. R~{s} R c { .~ ) < s i u 1 IV. 5 Re{§} > 0 11.474 Appendix A.49 . 4.48 V. with the shift tireorein: I c'"E(~) o 0 X ( S +a) = .= X(s)Re{s} < 0. S 1 theii use the shift theoreni: p"fE(f) 0.
3 .2 F ( s ) = ___ t+s+l n li? C H+t ss5 Solution 5.1 1 Solution 5.Appendix A. Solutions to Ilicl ExcmAses 475 Solution 5.
lim . Solutions i.s t Lz : I .1 = A1 (S t I)'(s + 4) + A2 + I ) ( s + 4) + A3 (S + 4) + l3 ( S i.o the>Exerciseh A2 = .476 Appeiidix A.T s3 : S" : 0= i l k + B 1 = 4A1+ =3 ill = N =  1 3 4A2 44A3 +R + A2 = 1 (14 4a1 .s2 = 10* (5 I.[ F ( s ) (*s 1 d l! s+f ds 4 ( + l)"]= 2 ( s +(s)+ 4)22 s .+ I 1 3 .arid 3 A3 = 1 can be c.B)= 1 . 2 s .1) S=l A1 = 1 2! lirn [F'(s) (s i siI (I" d.' <L lolls.4)2 (s 2 + 4)'  2(s+4 .iiig coefficients is tfieri possiblc with only two more eqw t.1)3 (. 9 ) 5=1 1 J8 2 ( s + 4)' 5=1 1 3 1 ( s + 1)2 1 (6 F(s) = . Eqiiat.dculated its in a).s 1 + t 1)3 1 +1 4 3 .4A3 .
j2 s i..~ 2 s+3 .j2 lirri [ F ( s )i s s 3 + 1+ j 2 ) ] = s + I+ .5 = IF' cos(2t) T L .s _ 3 _ . equating coefficients is trivia 1.72 A= sl+. 2 2 (S+ t j 2 23 + + + + + + + + + 0 f(1) o l u t i m 5.( s f l ) .4 F ( S ) = ___ 177 A A* + s + 1 + .1 .Appendix A.+ '1 2 2 Since thcrc are 110 poleb apart from the conjugated coinplcx pair.j 2  .' biii(2f)l ~ ( f ) . Solutioiis to the Exerc*iscs Solution 5.. As+B .ja  +2 s=1 j2 j4 1 .F(S)= s2 i.2s 5 32 2s 5 (s2 t 2s 1)44 (st1)* 22 ( S i l ) 2 F(s) = (s+1)2 3 ..
6 ( s $ l( ..E s 1 3 s 1 F(.. (but not a pole) into b’(%s).o calculate B is putting some value of this casc F’(s) is evaluated at s = 0.s2 + .1) 5 2 s s?.D +.1 2 2 ._____ 1/G + + ____ T l i e ~ : F ( s ) =I s2 s s+2 Ss.= 0 j B = 3 3 36 12 36 9 1/3 2/9 5/18 .s+3 R A trick t.3 2 .+ ++. 1 1 5 1 8 1 2 F(l) =.5 . .478 Auxtendix A.= .) ( s I.16s .s .3 ) )s 2 (s 1)2 + B c ‘ +s 41 ++s J.3 s1 Inverse trarisforrnation using Table 4.A F ( s )= __ ( s 41 ” 4.A ++F ( s )= sZ(s 2)(s+3)(.L ? .7 .+ . Solutions to the Exercises Solution 5.2) )s B .+ .B + .3 s s+2 5 4 R 1/3 5/18 1/6 s1 Determiiic B by evaluating F ( s ) at the zcro s = 1.v) =  1/3 .=1 2 hT1 2 F ( s ) = _____ (s+1)2 1 B 3 + ____ +.1 yicltls: ~ 13 / Solution 5.2 + c +.lu .2 s s Si3 4 . 2 D s +3 A = ( 5 + l)”F(s)( =.
2) f J . Soliitious to the Exercises 479 Inverse tri~risforrnnt.ionusing Table 4.Appcwlix A.1 yields: f ( t ) = ((1 .l + 3 P w 3 f ) E(t) .
]j(s+l.1 45Q) (s (s Ascertninmeiit of 12‘: H ( s ) = li s1 + 2) Y3 s3 + 4 9 I 30s + 52 .5 j ) ( s $.3 a) W ( s ) = I ! b) N ( 5 ) = K (5 (s1+1)(5 .5 i i i~ Fmrn Lhe polozero diagrani we bee that: 1 .l)(S J2)(S J2) + I{.4x0 hppeiiclix A.1 I 4’ d) H ( s ) = I< s 4 s(s + 3) ’ second order system Polezero diizgrmi o f the traiisfer function: Polezero diagram of the transfer function: ~ Q l i ~ t 6.1 . Sohitions to the Exercises Solution 6.J ) (.]j 11 52 2s t 2 s2+2s+2’ s2 s2 s c ~ ~ m d syst c r n order seconcl ordm system s1 .S+l+.
_______ (6  + 1)(.c{. t.(s 4.) + 1 A s t 1 +  B +JW() s B4 +.52 we deterrriiiir the system cliffe~ciitial(qiiat ioir aos + + 52 a) T l i c signal i s also ilheatly liiiown before i = 0. ]I s=2 =I (s + 1)Z 1  5=2 1 l =. so it is not.2y3 (s + l ) ( S + 2)2 j SE2 =1 Insertiiig thc coefficients gives: . S ( s ) with = I H(s) = 6 1 1’ R.s} > 1 do . Solintions to t hc Excrcises 481 FIQIXI II(s) 52s s3 t 4s2 +___.Apppridix A.s ( s+ l)(S .( s+2)2 ds [ d (s + l( )s 1 i 2)./U() +2)2 + s c 1 + s D 2 + ( S l?L 2 s c I ) < Finding the pxrtial fraction coeffiriciits: /I = . an initial xalw problen1.) It i s Y ( s ) I$(s) .++ d.
) .t .~ j w o ) + ( .e. 2) + + + 1 1 1 Oh5erve [hat for the inverse L a p l t ~ traiisform. i..1 i Re{s) < 0.w ~ ~ . the iegioii of converge i:. Solucioiis to the Exercises Y(s) = ~ 1+w$ 1 ( .182 wo ' Appendix A.~ J )..)(2w..s ts+1 (52 iu.j ) ( 5 . the r intcrsectioii of ROC{H} und RoC{X).+ ~ w ~ ~ ) (. ROC{Y) = {sl .w o .
l { + }. Solrrl ions to the Exerrises Putting in YI and 15.2 (go F 3 + 'pd Now t hc cornplete isolution can be given: = + external part internal part ~ o l ~ t i o7.2 513 25 + Iriwrse transfor uratiorr: Y(t) = P { H ( s ) X ( . 1 5t3 ] f ( t ) + 9" edt €(I) Conversion io cosiires H and B* in rrragnitiitlc and pliase * .0 [vo 1 5a3 10 5 1 = s + + s f 3 S2+1li + ___ [MO s+3  0 1 (using Eq. X(") 7 .(t) r lhtcrrnining c from the initial condition.(3) is oW.l ( S ) X ( S )= + . j RJ 16 1 Y ( s ) = H ( s ) .. y r ( t )= ~7('~') 483 +c ~('lt0) = 2 cos(4t .16) r) F'wtial fraction clec*oinposition: A l3 B* I .71 S . $ S f 3 S s .c + + 'Lcos(0) \4/.Apuendix A.d =12 c = yo y(t)  1.. ( [ ) ) =. 7.1.ained )liltion: g ( f ) = y)L(t) i/.3 Ii it) H ( s )= s 1 +3 b) X(h) 10.J4 3. y(0) = yo = 1/fL(0) U . s ) } L ..10 A = [ ( s 3 ) H j s ) X ( s ) ] / = .
with (4.1.1 from t>heblock diagram. and the initial coidit.1)) + + Y ( s ) = H ( s ) Xjs) t G(S)z ( 0 ) .2 e ?' exteriial part 1E l l ) internal part S o ~ ~ ~7.(t) 0 YY(. (7.53") .= [4(1 .16) (LL[SY(S) + O O Y ( S ) = p l i s X ( s ) .r(O+). 4 = 2zo () {/(OS)= ?J(O) I.F {2e P)3 .ions are not needed.)  y(0) .15): ~ r l y ( l ) ~ + i g ~ (=) t For nnilateral signals that start at t = 0.I s+ 2  120 + [.2 ) .15).~ s +1 X ( s ) . X(S) y(0)] Puttirig these pairs into (7.0.I)&(t. ( .0 .2. 0 .. Re{s} '> .______ S+l 1.34): ?. direct form 11 6 1 4 G ( S ) = . 2R H ( s ) =z _____ . y(t) = hppcndix A.C(E(1) .(t) 0.l ) s +1 I $ ( A T 2  4x0 .5 + 4 = 40 +2 b) Sirice the initial s t a t r has beeii given.3 f 'a4sd + 2 cos(41 .X(S) t ~ ~ ~ S O 2) 1 ~ 7.5 ~ Initial conditions from the block dingram: y(&) = . s ( t ) t 2p d'11] E(f  1) .r(0)]t I9(.X { ) ) P ' .484 B = I ele .z0 = il S i l = _ _ I _ + . (7. 5 .k ( t ) (t .+ pp i i 2 2 s I ___ P + s +1 2 g ( l ) . Solutioiis to the Exercises 4 o = arctan() 3 M in M 53" (!/(I P .+ ](P51) s sf 1 2 2  .4 n io Kq. yields (7.21) is suitable f i x t h e > solution.
because the stczies a w iiot the saIric: (7.Appendix A.@)  (s + l)(s + 3 ) 2 2(us + h) + (s + l( + 3) )s  b) No.58) oizly liolds for direct form IIT. Y ( S ) = H(s) . ~r 1 xs Solution 7.'Hospital: The impulse wspoiise (8. Solutioris to the Exercises .56) mtl (7. Solution 8.1 Tlic response of and RC lowpass filter to a square pulsc.7) i For TO 0 according to 1.2 t ~ O ~ ~ .(1 .57) are dilhrnt for ( ~ ~ choice o f s t R t C arid 41 niust ?x dctcririined beforfkmd.is: 1 T r E(t) S ~ l ~ 8. The relatitions (7. (8. X ( s ) t b) + ( s ~+( lG) (Ss+ 3 )  .4) with 1 ' h ( / )= e+ U .
t2) h3 Dc .ra(t) = ~ ( t )C(C .1 2 ' k 1 h(t).k ) = k=O (1)%(t .) q(t) = &(I) = = E(t) ' t . xi = k=O ( l)"(f  X) ii'(L(t) = A =..3 ~ o l u t 8.3 ~ ~ o a) . .k ) t I.o ( . .2 ) =.2) 2 .E { t .486 Appendix A. = c(t) ' qt) ' t t 2 ..+E(t) E(t 4) . 2 (0 ."E(1) 2 v k 1 t . Solutions to the Exercises c) 1 sr . .l ) % ( i.3 ' .1) t ~ (.
) = J . 1 '1 I [   $. dl.7 .r) d r = t r : y(t = f ( t . I: t y(t) y(t) = J(t 0  T)tlz = 4t .= Q .ion: z = (it.49). . .r) d r = 0 0.z(t) ?j4(4 * z:3(t) q ( t )* Q ( 1 ) g/3(t) = Q ( t ) * x:s(t) = 22(1) = = J. identical to (8. The xesiilt is.Q(t) q(t) yg(t) = q ( t j * Z:s(t) ys(t) = Q ( t ) * xg(t) * Z:i(l) * S o l ~ ~8.8 4 t Sollltion 8.l(t) yC. of course.ul(t> y. except that ~ ( z is 1nirrorr:d and ) shiftecl. d dr j =U d ci dt dr case 0 case 0 case 4 5 t < 4 : y ( t jj = fI ( t .Appendix A.YN(t) .s tl S ( l ) = &(LlLj dt dr Substitut. g ( t ) * X(j(t) y/7(t) = 5J(t)* z s ( t ) . 8. Solutions to the Exercises 4' 87 Solution 8.8~ o ~ t The calculation is perforrned as in Sec.(f.3.4.
= 0. lms gives the vdues R = 1 C : : 100. Imh. ROC(Lr(s)}mist also ' coritaiii the imaginary axis.U ] . I." l ~ ( f ) AI c o ~ ( w l+ pi) t 0.O.O1e(t)emlf c o s d l t with 01 = 0. 102 1 N ( s )= ICls~ 0.1 0. B = .01 s + 0.0 1s  A U $ h(t) = a ( t ) .1s + 1 + + 52 I s2 + k 0. Rightsided since h ( f )is the impulse rcsponstb of } a real system a d must therefore be cmzsal. Since ROC{ I ( s ) ) =KOC{N(s))nROe{vcs)}.488 Auueridix A.1  h) N ( s ) = 1 0 + 92 + 0. iinagmary axis.01 ancl (PI x 89" c) KCIC{/I} : R P { R> . L w'l 1.01s  1+ ~ (s + 01)2  0. The input signal can B.01.01s 4 0. Solufions to thtx Exercises Sorrnulizirig to 1 V. but mist d e t q 01 er Iiine.58. any 1 .1P or h ( t )= S ( l ) with Al M + e .LTIC amplitmk. d) Tlic system iesponsc ? ( t )convcrgcs if ROC{l(s) contains the.1 0. So crg > 0 and WO can be trcely chosen. I .005:d~ = fi33iiF M xI A0 and A = 0.
2 i o ~ ~ h) and c ) .la. since the regioii of convergence of the Laplace transform corvtaiiis the imaginary axis.7): 1 F{E(f)} = nd(w).Nevertheless.17): The dwlity princple is used as follows: S(t.r(h)) does riot co~iveige.S ~ C ( r ( f ) ) = L{€(t 5) . + S . 01 1 pair (9.&(t 5 ) ) = 5 finitr duration. Solutions to the Exercises 381) t5s (+ . d and e do not converge to a function.. for a) TJre t h p niodulatiori theorem t m pair (9.3 The Fourier integrals from Exertisr 9. sirwe .c(t) has L(J(t))= L(G(f)) = I 4 I 2..7): for (1) 1Jsc the similarity theorcm on pair (9. 30 c. .)wr U().? for e ) Use the yriiiciple of duality F{eJW"E) = 2 7 d ( W + "0). the Fouricr transform clocs exisL in tlic form of a dist ri hut ion.Appendix A. s E C. Re{s) < 0 see e) F{. 2nS(LL' .) . ~ o l ~ 9.z> = 2nb(w + ?) Wliere ir is any constant that can be appopriately replaced in thr1 result by . s EC ejWf d l (1) FF(x(t))  1 rxl 7c 5tt) cJio' d t = 1 0 cc does not converge * C{T(f)) = b 1 . (Lying on tlw border is not srtllicient!) Solution 9.I . a) * LC(i(f)) does not convergc 'Calciilitting the Fourier integral does not provide the soliit ion betanse the integral will not cwivergt~.
.a) Zwos of si(x) at 1ww: won 47r ct = n?. = 47r 1 2 c) rect (207r "1 .* I * 871. n7r n E Z\ (01. 6: 1 + w'o = 4 (3 h) Only calculate the aiea of thc triangle in Fig. cxI 7 5 ( t )dt = .
real part evel1.he Exercises 49 1 If 11 = 0. Solutioiis t.no phase has been defined.t for (1 < t < 4 0 otherwise olulion 9..?U) = . irr~agirrarv part odd: X(JW) = X"(gc*?) b) Real part odd.e.Anuendix A.X ' ( J U ) .4 < t < 0 4 . write riidl x XI (s) =  s2+2si17 5s +5 ~ (Sf 5(s i 1) 1 ) Z { 42 with Re{s} > 1 1 4tt for .8 a) Conjugate symmetry. i. iniaginary part even: X( .o t.
+ .  . + __ +=E. / w ) . .. = yJt) = + ..  1 To produce the desired clrange of sign (+ .. ell forming the complex coirjugatc (++): Y(.. . Both imaginary parts chnngc sign: (JW) =X * (j.' /  . .11..  ....jw) iiitist = + = + ... . . .492 Appendix A.+).+.) a.... the sigri of the argument t~c rcvcrscd (+. Solutions tothe Exercises Yh(t) l5(..  ......(pd) X * ( . I .*..  .
4: . 1 Jt+ U 1 t 2m(~)r~"~ 2TJ E( +)paw .o  + It b) fxoiri sec.n ( W WT)] Solution 9.o. 9.3: c(t)e'" o Ihiality .o1 Solution 9.2.2.]CC 1 c) from sec.CS(W . c ( t ) 0 1 1)uaiity: rii(f) . . the rules for t3erimtcd clclta iinpiilsrs (8.12 a) from sec. . 9.W T ) ] Y ( p ) = [.23) arc used.(Y.A u m d i x A.: * / X ( J W ) i 27rnzd(u))] X ( ] ( W WT))]  "X(J(W 2 +WT)) + 77r?7t[fi(W f u r ).11 77r[6(W tW T ) .13 i! s . To calcrilate thc integral. Soluhions t o the Exercises 493 Solution 9. 9.4. 27r = 1 .
.Y)] 0 I 1 olution 10.1 a) 20 log 10 =.he power ratio. Solution 10. It gives tile sariie resiilt as above.n io iUote: ?'he formnla I = 10 log c m also bc used to calculate the dmplifiration / Kut directly Gom (.1% a) dx (t)  dt = = [E(f 1 T 1 +T )  'LE(t) t E(t  Tj] d". Solutions to the Exercise8 Solution 9.3 ~~agxiitmie = 20 log I I T ( J W ) I = GO CIB20 log v Jm ci13. p.. . 20 d R b) 80 dB r) 3 ill3 d) 34 dR c ) 6 dB S o ~ ~ ~ t 10.494 Appendix A.c(t) df 2 [6(t 1' 0 *T ) 0 I  2S(1) 0 + b(t .
20 log II in dB H I 3 10 100 IK * 0 Plisse sketch: Proceed as in Sec. 10.J 0 84" 10k jo. "Sg { fI} 0.1).3 .) 20 6" 5J 5 17 45O lk 0. 10.2.4.Appendix A.01 40 90" 20 log 102 = 40 dfS ad: with Amplitude sketch: Proceed real poles arid zeros. Solutions to the Exercises 495 w H(S) r/[dB] 20 0 1 10 10 100 9: .1 20 90" J look 0. (Sec.l 1 10 100 1K 0" 45" 900 §o~~t~on 10.4.1 .5 .
500 = ) (I+ s 4.1 = a( + l ) s 1 s + 1000 . Solutions to the Exrrciscas H(s) = .X ( S )) Y S  1 .496 Appendix A.
8 a) H ( ~= 1 0 4 .1) . so 6 dB in this C.0206 dB 2 This value is roimdcd in tlir Rode plot to 60 dB lmtling to a dcviation of 0 '3433%. ('5 i (s + 10)(s+ l(iK) 0.(5. (jl0+ f o p 1 (310'2 x 14.01% c) 3 d U pcr polc. ) (9 (s + + 1)(s + 0. Solur ioiis to the Exercises 497 ncviatiori from Bode plot: 0.Appeiitlix A .1) l)(S + IO)(S t 1000) = 104 . Exact calcrilation: IN(jlO)/= 1. Solution 10.LSP.
. 0 Phase sketch: arg { fi) 100 ' 1000 = 10^20 dH The system has a bandpass behaviour. .1.1 and 20 dB rcspwtively. to the Exercises A. but for w < 0. 1)) Arriplitnde sketch: 20 log /HI 4 jH(g L00)l Fz 1 ' . < [Ifremains constant Tlir ck voltage.10' it has dcsccnded to of thc maximum ( q i i \ a l r n t to 20 dB).498 k~ppc~ntiix Solutionr. and at i~ =. l l r e niaxirriiiiii aniphtic*stion i h 10. & c) The response to ~ ( tfor I ?' oc corresponds to the dc voltage gain. i. The frequency cl = 0 is not in the diagram. At U! = 1. gaiirs arc 0. ) IH(w = 0)l.e>.
Dorrbli~rg thc frequency qriath iiplcs tbc\ amplificatioii.12 ~o~ P) 40 ~ dB dec* dB t l ) n . w h m ii: << 100. 20 der The amplification inrreases exporieritially at 40%.Appendix A.ions to the Kxerrises I____ 499 dB dB a) 20 b) 20 <lW dec S o l ~ t 10. Solut. increasing it by 20 log 4 = 12 dB. .
.Thc amplitude skctch iiirlicates ulq = 100.
Solutions to the Esrrriscs 50 1 Solution 10.Appendix A . 20 log 1 HI + 1 1 bo .15 a) In this question only thc inaxiniuiii amplitude can be stated The values at w = I and U? = 100 brlor1g t o c).
In the real world this corresponds t o the a lossfree circuit. Solutions t.3 .3 Solution 11.502 Appendix .o the Excrciscs ‘ I: E corner frequencies 1 n The condition is a = 0. Solution 11.4.
0 I .Appendix A. Soliitions to the Exrrrises 503 to = 1 2: Solution 11.4 0 .
The period of the sLiirirrirtl signe.5 Isirl((1 . Sohitions to the Exercises Solution 11.he sum of licrioclic hiiictioiis is only pcriodic if cnch period is a rat.6 a.he otliers. 2.ioiial proportiori of t. aiid the fuiidn. it): fiindnmeiital freqiieiiry of circnit 3u?0.al frequencies.5 Solution 11.llrli tlic snisllest common inultiplc of the individual periods.t)] .504 Apprridis A.f i ) i ~ o t ) sin((1 + + JT)q. t.hc individual fuiiclament.l 2' is t.menta1 frequency of thc signal f = is t. period siii(dOtj cos(&wot) = Yi = x2(t) 0.) I n grncral.hr largest coimnion divisor of t.
Apperidix A .7 =0 hr p # 1 = 0 for p +0 = (1 for 11 # 1 sinre integration over' 17. periods of e l W l ~ equals z ~ r o ' . Solutions to the Exercises 505 Solution 11.
Solutions t o the Exercises Solution 11.9 a) Z(t) 0 = si(7rt) .elTop'" ( h c a u s r of t h e selectivc property of thc inipulsc at w = 0) Solution 11.8 X(:pj = iLL (T )= 27r W 77 Sw ( k  27rk) Fourier series: X(JW) = P A.si(7rt) I .506 Appendix A..
u rriultiplie(I ~r)ya facttor $ = . Solution 11. With (11.18) t h e re' quircd relationship follows ..c(t).11 a ) Converting (11.14) yields where X ~ ( J W is) the spectrum of one period of .The weights of the dclta impulses corrcsponds to the sample values of X ( J W ' ) at frequencies w = .
14 For a. cyclic convolution t.13 Solution 11. Solutioris to the Exciascs  ~ 4 7r k 1 (2k 1 1 )2 S(w ..508 Appmdix A.o exist. h t h sigrials iriiist have the sallie period: .27r(2k + 1)) Solution 11.
from Ex. Svlulioris to the Exer rises 509 i i Using the Foiiriei srrics for z ( t ) .hppeiidix A. Solution 11.15 .11.13.
Aliasing in case 3 .. L I 5 6 "Ig .. L 0 otlirrwise . Critical sampling in case 2 Solution 11.5 1 2 3 4 O@ case 4: wag =  27r = Jic! ' Y T 4 c ~ 1 2 3 4 . .rert 27r 271 dg 1+U for . Sohitions to the Exercises / 0..16 27r 1 dg 27r X ( p ) = .510 Appendix A.d g < u i < 0 forO<wIw.
Solutions t o the ExPrrises 51 1 Solution 11.17 p 4'b t t .Apnendis A.
CO.19 a) L .512 Appendix A.Solutioris to the bkercises 3. 71i 9li to Solution 11.
liebaschand. T/. S O (I1 42) ii \atished.r must 2 rcniove all imagrs outside of the baseband. baseband baseband . berauw the allocatcrl frcqueiicy lmid is a riiiiltiplr of the bantltvirltli. t IN 0 No a. so n. a i d 1 d the basrl)and p a s unchanged H(jw) j l'wo possilile solutions w 1 = aiid d 4 = .20 a) r w l handpass signal h) Yes.liasirig (overlapping imagcs) may o(~iir The recoiistructiorr h1tt. 4 1 Solution 11.ii). 2 criteria inlist be satisfied: Ail iiuagc of the spectrurri Y ( J u ) iiiust lie exactly 111 1. Solutions to the Escrcises 513 Wlieii choosing ~34.Appciidix A.a = d o .
5 14 .. 2 ~ ~ 2 (a) e ) a.mln =? n l = 24).mln 2Aw = 2uo = d) no. = 2Au 13wo yes.. (11.= ~ A ( JWith 2l 7 T uo = 0 and the above condit...ica1sampling is alwqs possible.1 = 3 .iicl f ) are complex bardpitss sigiiitls =+ crit..23 .4pperidix A.. Solutions to the Exercises Solution 11. L)..44) becomes: Solution 1~1. l L Solution 11.ion. U. 8 ~ ~ 1 11) yes. ua... u~!.42) a) no..the following holds for critical sampling of real bandpass signals: = .. In both cases. : . iL'rr..1..1..22 According to Ta.21 a) t o d) are real bantlpass signals =+ check (11..ble 11..
+ SO= 72 H a . = 7r  term in square hrackcts arc at 24 Hz and 48 Hz. r.Appendix A. the zcros o the f . Solutions t o the Exercises 515 h) A(j0) = 7l  c) 7'1ie first zcro of the aperlure furicfion should be at half the saiiiple freqiiency.ion should be equal to 7r. The first zero of'the si trrrii IS at The ros term has peiiotl U+. the argument ol Ihe si fiinct. Accordingly.
) 1 4 t cos(l0nf) s1n(107rt) (i) LIP.516 Appeiidix A. ( I = 7r 2 5 . Soliit ions to thc Excrcises Solution 12.1 b) x n ( t )= (.
a= 0 R =0 ? l Is) c = ln0.3 7i .Appendix A. . (l:?l R2 cl)C=O Solution 12. Solutioiis 10 the Exeicises 517 Solution 12. c) C=lnO.2 a) :? = 2 .Y .9 .
] h) when k =  t T X(c'") = X.lie result.irz) = I" A I L ~ J y l b / "with * f uiidanieiital period .518 Armmtlix A. the spectrum of a serics usually i s written x ( e j ' ) . appear iii t. c) Fourier series X(e.(jw) mit R = wT Note: Even wheii e j n ternis do not.[I. Solutions to the b:xPrcisrs r[k] 277 ' = ..
callpd 2' Solution 12. H2 is a Iiighpa. T = . Solution 12.6 a) It is cvidenl' h l ? l spectrum X(gw) that X ( J W ' ) 0 for = 7 r a t the nycluist rate.ss filter.miist be choseri. Solutioiis lo the Exercises 519 This Fouricr series correspoiicls to the first step of a ) . d) A discrrtc lowpass with cutoff freyueiiry .5 7l R halfhaiid lowpass.To sample WLl . l ~ j l > u7.Appendix A.
X ( ej") f I + Solution 12.13). Solutions to thc Excrcises b) Accaording t'o (12.ilit..t'he Flransform of a sampled signal corresponcls to the F*trmsform of tjhe ccluivalent.7 Without loss of gcner. discreLe series for R = w7'.y we choose 0 0 F [n: 7r] .520 Appeiidix A.
1 .8 k k Solution 12.Appeiidix A . Solutions t>ot hc Excrriscs 521 Solution 12.10 Inverse Lraiisfomi according to (12.9 Solution 12. fbl Solution 13.16): v I hfodulation thcortwi = d k l .
Xs(z) = ___ 2a .1n ’ a < 12) <U 1 only exists whcn (1 <1 Solution 13.z 0.8 < J i< 1.0. .2 &(z) = z ...8 1 CI X q ( z ) does not cxist because the regions of‘ convergence clo not overlay.88 ’ 7 0.5 < 1 ~ < (1. __ 1 ..Q I .82 k=O ~ ___ _  1.x. 0. Solutions to thr Exercisrs 10 k10 h=O 10 10 h =o 1 .522 0 Appenciix A.[k] = trkclk  I] + c1 “[k]O.3 .25 / Solution 13.5 ~ 0. z0.
.AiiDcndix A.ritk All filters arc lowpass filters.4 X2(z)=zl+1+a= z”zi1 1 X.(z) = z2 + 21 + 1 _ i z z + +I _ 2 ~ X2‘k1 ~. Solutions to the Exercises 523 Solution 13. .
5 .C[k x[k h h Solution 13.6 9.7 Xi..)= 0: 1 x.524 Arivrlitlix A.5 x . ROC . Then it follows wlth > 0. Y a.) Z{. Solutions to tlic Exrrciscs Solution 13.) = 2 { r [ k ] } ..= __y (z) ( Z T Solution 13.
0.Z 1) 5. 3 . O < 22 0. I No poles except at iiiliiiity + finite. 1 .ial fractions of : is carried out by sp1it)ting X.5).10 R) H ..Appendix A.I&(?) = infinite. &(?) = + finite. ROC . the iiiverse transform p. 13.+ j  +I .rt. :)( ~ + . ROC . f) .5 2(.8 The locntioii of the poles is the dccidiiig factor (see Sec. ( Z ) = ( : .1 > 0.9 Because the series a. No poles r x ( y ) t at the origin + finite.i.~)’E[~] Solution 13. KO(‘ : 121 < K 2 .~ ) ( Z ’ ~ +1)=(:I) = ( .(. c  r2[k] = 46[kj + (4  GX)(O.) .5 4.rc infinite. Solulioiis to the Exercises 525 Solution 13.3 #? . ROC : O < 1 ~ 1 1 ~ 1 + * 7 ~ +  ~ Solution 13.
12 a) .r‘ .21 + with the shift theorem: X ( 2 ) = 2 { 5 [ k ] } A (1.t z : j zrL= e 2 I. and a constant factor Solution 13. .r.2 ~ [ . . .1 ] k 1 +~ [ k2. n = 0.z2’’1 = 21 p 7 . . Solutions to the Exercises The diffcrrncc is a shift.[k] = ~ [ k. pole witallnmltiplicity of 2r.r 1 i s riullificd by a. + 1al y? =0 . 1. . 7/ 7.2 b) double zeros at z7’+l= 1 ‘rhc single pole a.526 Appendix A .r.2. zero at z = 1.11 .
””kl + aJ:zlkl + + . because S{:r[k. not. Tl..Appeiidix A . Solutions to the Exercises 527 k g) L.]) S { Q [ k .Nj = 11) L L I ..] + S{r1[l.N ] } = :r[p. TI. becausr difference equatioiis describe LTI systems. bemuse S { x l [ k ] ~ [ k ] } = u ~ ~ ~ ~ ]# ” ~ [ ~ . ] ) = a. not I. i) I. see d) j ) TI.
?/[A.re invariant for a cert.= 0 2 .sii:i~flziivnrznn.ic .: ] = x[2(k. .). a. C)  N] : s [ 2 k .rni is iiivariailt.iarit.I] .ed by an even iiiirriber of cyclcs.N)] y[k V} # b) likewise shihvariant. Note: Systcms t h a t .3 a ) lJ[1] = 0 y'O] = Y[O] = 1 U [ I j = . Solutiorii to the Exercisrs Solution 14.N] 'L'lir systciii sets to zero every sample with odd k : Exaiuplc for N = 2: shiftrnvar.t.528 Appendix A. because S { S [ k .2 ' 1 = 2 1 y[2. L [ 1 ] 2y[k . = x[2] 2y[k . The syst.1 = 0 .2 a) shiftvasiaiit.ain shift m d its multiples are called perzod.q[k].1 = 3  "I  . 11ut shifting: x [ k ]by an odd niimber of cycles does iiol deliver a shifted vrrsiori of' . (2) . see a. whcn shift. Solution 14.
vrxc [OJ) = 3.y[3] = 5 Solution 14.g[1.25 + 0. DF I: Begin wit. (1 y.1 .+[ I ] = 2 1 + 0.a]= 4 2 1 ~ y[O] = y 3 = 431 .+[O] =2 1 =2 .+[3] = 2 . 1 + . ( 1 + 3 .rioriic:.li the fcedback brmch.y [ l ] = 6 1 ] 0 4 1*[1] ~ [ 2 = 0 [] y[S] = .tlfolms. Solution 14.5 .“5] .Apperidix A . Solutioris to t.23) = 1.4 y[k] = J [ k ].  21 1 !).5 . . 5 ) = 4. .6 a) 0 h iiiiique assigizmcnt of iiiii. 3 .2 ) y[4] 0 2 .2 ycxt[2] = 2 ’ 1 + 0. 2 > .. (4) 3 ’  + + = 4 : 5 g[k] ( 1)A ( k 4l)&!k] b) y[k] = 1..ial states is orily possible with ca.5 .lir Exerriscs 529 y[3] = 0 .. ( I + 4.625 Iriilial condition Solution 14.2 .3 .2 3 ( .5 Intcrrial part: 1eaxc oiit iiiput S i g d External part: sct initial state> t o zero y.
. .I ] ) I] = 2x[k)+ 0.ial condiiion is fulfilled by 0 . 5 [O] O.52[0]] =+ . 2. R.rrmirie t.y/int[O]. 14. Solutioiis t o thr Exercises .x[O] = 3.0. 5 ~ [0] 1I = .tiori by 2 yields the coefficients from a).t[0.I] + Differerice equation: 9 [ k ]. 5 ~[O] 0 . (11) ' (I) in (11) : z [ k + 1 = z [ k ] y[k] 1 yik] yjk] = = 2X[[k:] + + 0. yint[O] = Initial condition: 0 z[O]= 3[0] * z[O]= DF ITI: scc also Fig. 5j/[k  + lj[k .5 'yiI.[ k] x  'l'hc init. 14. 0 DF 11.5s\k I]  MultipIying the difference rqua.l 2:c[k] O.7 ancl Fig.r[k .S(. = 0.9.401 = 2 b) (I) .S shown in Fig. and S o t e : Thc internal part alone is suficierit to det.54k.530 Appeiidix A.r).r)z2[0! ~ ~ 0 . 1 2 + + +4 O.he initial states.
t [0] = g[O] . x* . Oiily then will tlic auxiliary cwditions bc fulfilled.yext[O] = . h c r c i s e 14.ll. SPP Solution 14..9 a) IIR.8 Solution 14. system.lie output signal is fed bad< int.I a n d yillt [I] = = y [ l ] .he system.4.Appendix A. which means that t. _ .7 531 DF I1 DF 111 2 U =1 [kl The initial states realise yi.)(r q ~~ A P 7 7  + ~A* _ . Solutions Lo the Exprcisrs Solution 14.o t. because 131 ( 2 ) has poles outside Lhe origin.  HI (2) 2  UZ ( 2 .yrxt[l] = 6.
.* 0.11 External part: .5 ' > 0..5:l 1 .25. 2 4 4 H'L(?) = a .I + 0.10 aj H ( z j = 2 D..5   0.522 . c c .552 Appcndix . (1 z . .2525) > +  Solution 14.5 X(2j = 2. z1 /i/ > 1 Solution 14.0. Solutions to the Excrcises h) Calculatr the first five values of h l [ k ] and use therri for h2[k]: a a a =+ h z [ k ]= f f .52' x + ~ 22 z + 0. ..0.4.0.0.
Amendix A.12 .eriial part: Solution 14. Solutions to tlie b!xerciscs 53 3 Irit.
1 .534 Appendix A.14 Solution 15.13 Solution 14. Solutions to t IIP flxcrciseh Solution 14.
S(w 1W O ) ] 'Ft{coswc... j ~ [ ? i ( d  . 2r.2U") ~ h(u + 2iJ. in the linirdomain by 2 Solution 15.I)] T  Z { c o s 2 w o t } = siri2wot = cos In all cases! the Hilbert trmsforrn caiises a phase shift.3 a) h ( t ) = .GO) + S(w' + w O ) ].o 0 H (.  1 . "reacts" to future x [ k ] wheii k > O e.p) = gsign(w) 7t T 1 . g[l] . Solutions to thc Exewiscs a) not cau5d 535 b) caausxl c) causal (1) causal e ) not causal.ApDenclix A.d(w' 2 0 .g.WO) . (~jsign(a))= .t} = sin wof = cos d) 'H{cos2 U ( ] t }00 ..[a] . p [ b ( w.
vie obtain: oc ‘~t{x.trrmg Yy. Solutions to the Exercises The iiiagriitiide of X ( J W ) stays unchanged until d = 0: y(t) ist real. Y(jc. and therefore 1X(1. according to (9.F(t)= } [a.94): m X ( p ) = X * ( .61 and Y ( .) has conjiigate symmctry.536 Appmdix A.renl o ~ m ~ ~ q +a1u.imag ..1)12 is an cvcn function The Hilbert transforin of a real sigilal creates an orfhogonnl sigiial With the result froin Exercise 15 2 and sign(0) = 0. Using the symmetry scheme 9.rm~ 0 Xu. w ~ x r 7l +  ~ 0 ~yu.j s i g n ( w ) X ( p ) . l w )= . Since is y ( t ) real. sin(duvt) 6 .r(:iLi 00 yg?reai Tlie even p r t of z(t) becomes odd by the Hilbert trmsform and vice versa. since both :z.rrai lnld xzi. we obt. cos(cuovt)j ./ J ) .(t)and h ( l ) are reel.ain: Zy.
Appendix A. L$t) 7r . . Solutions tlo t.s1( W'g . J 0 ?(I) = .he Exercisw 53'7 .
and wing Parseval’s equation.538 AI>pcndix A. s Solution 15. sigiials with the same magnitude spectlum liavc the same energy. ic! = I) d o ~ not.i?(t) 0 = E{z(l)) 1 j. i w ) + ~ X . ~ ( . The exception at.ra.mof a.nsfor. Solutions to the Exercises Solution 15.r(jd) : jsign(w). ( ~ ~ t/w # )I I) According LCI Parseval.6 13) The Foiir~ert. make any difference after integration.s follows: 0 . Kinowing that. ixal function IS even. ( p ) l =I X . the energy ol (t) arid x * ( t ) be calculaled i\.5 .
Solutioiis 10 the Exrrcises 539 x w E1 El Solution 15..7 a) For a cansal system : 1 1 &(p)P(JLd) * = 77W = 8{P(p)} 1 'lhis J iclds: 1 H(gw) = P ( p ) .Apgciidix A .J .F J ( I / W ) b i7 *w = p ( t ) + p(t)sign(t) 2p(t)c(t) = .
8 S E I \ 1 ( j ~ ): 2 .Jw) =1 bandwidth S(. t h c qminetry properties (9. is p(L) is real a i d cvcn q ( f ) i s iinagiiiary arid odd Q ( ~ L J ) I P i d and odd 1s Solution 15. Solutioris to [.jw) .lie bkrrcises Siiicc P ( J w ) real aiid even. w g = 2~ 8 k H 7 5 bandwidth (.61) show that.540 Appendix A.
he Exercises 541 Solution 16.0 . l z t . we oht aiii 57i 5zr .i.[k]I< A l l .2 M 3 stable Bouiidd input: h=x SiiffiriPiit I. + + ~ Poles lie in the left halfplaiic Solution 16.1 a r i d the riiodiilation theorem of the L a p l x r traiisforiii.H ( s )= .1)2 (57r)J (s .1 = 10 j stable I)) with Table 4./5n ( s 0.s p ) ( s s. s p = . :\I1 <w for bounded oiitput: .Appendix A.. Solutioris to t.
e." decays where k  00.1 ?s2 + as + I (s ~ ty noriiial c) All Leios inside Ihe unit sample are j rriiniiriuiri pliahe..lie iiiiit c:irc*le. jzlp) = i3) 1. if I. H ( z =.5 The iiitrrrial part with order z can be split into paitial fractions: ?'he series rirclr.I ) is finite 2. 121 < Zeros: zI l 2 = 0 Poles: z l p = $(I * j f i ) . 2. ) 1. Numerator is a Hurcvitz poly z 1 2 +4 stable. if thc polc lies w i t h the unit . d) Yes.riti stable for la/> I ! because the poles of a leftbsirled sequence have to lie outside of t. z2  < 1 + stable. Possible regions of convergelice: ) ( 0 1 + H ~ ( z is iiot. Solution 16.I < 1.. 2. . /?I > ( U + H ~ ( zis causal and stable for iuI < 1.532 Appendix A. i. causal a. .Solutions to the Exercises H l ( z ) is stable for all a. evidciit from the diffeiwce equatiun.
u4 = =+ N ( s ) is a Hurwita polyriorriial. k)eraLiw tlir denoiiiiriator of H ( s ) i\ iiot a Hurwitz polyriomid.5 8 ' ~ 7>0 325 > 0 a:< alo.s in N ( s ) = .or polynomial N ( s ) .hc cler splane + lntersect~oii lef(.10.cstj: a. becaiise no zeros lie iii the right hdfplane. .o the Exercises Solution 16. Hurwit.erim H ( s ) ha.(1'.= ala.6 a) Derioniiria1.= a1 = 4. because zcros lie in the right halfplane. .%" t8s' t 7s i 2 1 + possibly stablc. because zeros lie in the right halfplane. H ( s ) does not have mininiuni phase. all cocfficients in = .fikient.5[s" all coeficierits are positive a. circle on the splane.8 H (s) i s unstable. = 162. : .qs) + + 110 t.< = a.5 > 0 A. Hurwitz polv~io~~iial.z t. Solut.sR s 2 are positfive since there is with s2 no1 il Hurwit#zpoiynorriial.ions t. so H ( s ) is stable.nd exist 543 + /I.:( = 4.s iniiiiniuni pliase.U.7 Since tlir transfoim is t h r same in both directioiis: 0 Inside the unit circle of the 2plaaic + left half of tlie splane 0 lelt half of the zpla. circle on the zplaue Iiitersect. Solution 16.5s" 8s2 7s + 2 are positjive 3 so it cariiiot be a. half of tlir unit. 0.ioii A lcft half of llie unit. + .2  o. Solution 16.5 > 0  u.nc + inside tlie unit cirrle of t. H ( s ) is not stable.2. arid ~ ~ + H (s) does not have minimum phase.a3uL~ f & .s4 4.n. c ) Kot.Appendix A.ll cot. so H ( s ) is not stable. 1 ) ) Not a.
10 a) . ( t ) is intrgrablc i stahlP Solution 16. Solutions to the Exercises l.(t) = sin(Gt) = u(t) Solution 16.9 a) H ( s ) has poles at s = f j + unstable c) s2  Xs + 1 must be a Hurwitz po1yriomi.543 Appendix A.tl + K < (1 h.
1 Solution 16. F(s) + ~~ _ s2  1 2s +5 + K . sp2 f 0 Stablisatian succeeds if 2 < 1 < m.Appendix A.11 E(s) a) II(S) = ~ _ 1 K .= 2 + 2 fi : s p . Solutioiis to the ExerCisCs points for root locus: r !d<J K =0 : h= 2 sp = 1 1 3 2 tsTi &J =  & f K=4: s p = 1 & J 2 double pole I<.& I< + oc' : hpl 02 .
verages are different. .z  # j i Ensernl. Ensemble 1: Yes Enseniblc 2: No: because the erisemble rueail squared E{:i?((t)} i s timcdependent.2)s + 5 necessary + suficierit t ha. this case the timeaverages squared : r L z ( t are also difin ) fereiit. The ensemble mean sqiiared E{:? ( t ) )IS c c p d to t81ictimemcrage squared z i 2 ( t ) for e. b) For weak ergodic processes: .nsernbles 1 and 2: Yes _ _ Ensemble 3: No. 7 s2 IIurwitx + stabilisation for li :> 2 (compare Exercise 16. hiit different smplitudcs and therefore different mean powers x t 2( I ) .) According t.. Solution 17. h) II(s) = 3.( t Only ensembles 3 and 4 can belong t.546 Apprndix A. (xt:i(t)# s j ( t ) . Ensemble 5: no.t a11 coefficierit.:~  (t) = constant. E.k stationary raridom processes:  p:{.veragcs squared x.t 2: the ACF ~ ~ t ~ t)does . ~ Ensemble 3: No: hecaiise the tiimca.11 sqi.i2( t ) of each sample function are different.: except a.ilueaverage xi(t) of ca.o weak st.de 4.at.rcd E{x'((t)} is t. a.o the linear t.10) + ( K . halfplane for any k . The sample functions hamr the smne mem. = .imedependent. Enscniblc 11: KO. because the ciiscmblc 1mea. Yes Hrisemblc 5 : No. E{s(L)}# constant.o (1 7. because llie eriserrible mcnn is timedependent.ia.sare positive Solution 17.15).1 5.1 The linear cnseinblc mean E{x(t)} is equal t.. at 1. Solutions to the b:xcvAsrs Cannot be st#abilisedbece.ioiiary processes. for we.iisc both poles do riot lie in thc lcft.ch sample function.(t) constant: except.2 a. becaxsc the timea.D.14) und [17.i.xh sample function.not depend on t: except.
This does not.20). E{z(t)) = 0.. $ ) ) ~ } E ( [ s ( i j y ( t ).igcsof any sample funct.ndom process is not. e. Wit.e :c. The expected value at a c.imeaverag.liies of the output signal 2( t ) . .he ramloin proc:esses are weak ergodic.most by as many samples its there are on the CD).his ra. In Exercise 17.. Applies to cnseniblcs 1. =+ t.3 a) Linear t.is spread across many poirii.s of the CD signal (at. ( t l ). and therefore: C. The condition only applies to eriseiiible 1.Q i.ain point t o 3 F:(z(to)}.g. Since music signals in general do not h a w dc components. h) With first order expectred vrtliies wc can only discuss the condition for stationnrity given in (17.lie random process is not ergodic. t deterininistic sigiial docs not change the variance. however. K(. ( l ) }= 7 + + =O E{y”t)} /I. lbllowing t. l a . x .4 i.o the loudness of the music.2 ( t ) .14): In a) we xerta.y(t). t ) ( 0. j none o ’t.Appendix A .(LL&) tP +L~Lly(t))]2) Since y ( t ) is deterministalc.ion. is trrlce~nfrom different points on the CD The a\iera. because 111 our experiment. =3  S o l u t i o n 17. r . = E { ( z ( t ) / l . = P.p u ( t )= .5 IT = ~ { ( v ( . because it is determined hrrl rnaiiy va.lb thc special case t l = t 2 is investigated.tionary. ( 1 2 ) ..r2 ( t ) }= x. f Solution 17.liere is iio tic cornporient. & ) ) 2 } T  + = D: = 10 Adding .he corresponding tsiineaver. it cannot be weak ergodic.21). weak stn.) . r 2 = 2 fiW(t)} = E((:x:.li thc same notation we can assunie E(x2(t)} = constant. S o l u t i o n 17. which for different sample fiiiictions 7. __ c) For ergodic:itjy:thc first and second order eiiseinblc nieaiis rnusl agree with t.ges are generally different. hewuse of (17.   E { : c ( t l ) ~ ( 1 2 ) )= z . r .iidom process could be stationary.P . = E{s(l)} = constant.ined that p.: = + S 2 ( t ) ) 2 ) = E ( r ? ( t ) } 2 E{21(l):x2(t)} E { .(t)= O! because t. 2 and 4: see Exercise 1 7 . Solutions to the Exercises 547  E{s(t)} = z i ( t ) . + the ra.ert. Becaiise t. h sample function conies from it ritiidom time. apply because to form thc timeaverage sqiiared xi2(t)only a ten second section of the CD i s considered.because of (17.
548 Ap.o thc F.yeridix . Solutioris t.6 .4.xcr cises Solution 17.
eh n i t h r whole CD are virl ually ulirhnged wit.!.r(l+ r ) x * ( t ) = E { K .r(t[.Appendix A. the ACF p.o Excrcise 17. 10s) = } + . 91 G = 15.rgodir. Exercise 17.~ = 3.ification is siniilar t.l. I)ec.11 As every sample flinction has a period of IOs. c:liniige with t..1 = E{..y"[Ic]} .}= .1) E { 2[ k ] } = 1 + 4 + 9 + 1 ( i + 2 5 + 3 6 6 35 Q = E { x ~ [ ~p.< + to I).(t.ion is given by tp.[X] =6 1' % E Z3agrre wit.. because t. A good apprrxirnat.hiii 10s: p.8 a) E.17 b) Likewise rrgodir.3t() = qr.to) E { z 2 ( l n )N corist.nnt.. N t Z./a. I<*} = p p r } Solution 17.aiise tlie properlies of a (lie do not.g2 2 =0 Neither stationary nor ergodir:. For k: = 3 N .8a).l.:.= 2.3: Tlle correlation propert'ies of a.c.(t(1.] from E{$[k]} = 36 ay [A.5 ( s w Example 17.r(ti.. sigal inixt.92 : ] 12 ~ .() = E{.lw ensemble means are riot constmt Solution 17. because tlie prupcrties of a.t~.~. '1%~ just..= 1517 The ensemble rtiea~isof y[kj for k # 3 N .ur.li tliosr of XI/. tlie do not chnngc with time. p.10 pr. 11. 1s also 1z ) pcriorl in t with period of 10s. indepcndent of lime / o .irnc p:. Solutions to t h c F:xrrc*isrs 5 19 Solution 17.
550 ADuenclix A.j62 ser derivat’ioii of ( 17.tcd: m .?.. b) ‘ 1 ’ 2 ~spccial case of r = 0 i n (17. Power: lirri  Power of t’he ac coiiiponenl: (d(L) .16 For t 4c a.(t z)w(t)} (17.13 LVith dctcriniriislic signal. Solutions to the Exercises Solution 17.!i. t h e quantities in qurstioii must IF calculated with tiiueaverages.48).vL:!.. ~ ~ .$o)z Solution 17./i.o ( 17.12 a) E{(. = h’{.nE{y(t 4 ~ p y ) >=  T)} .15 .( = E{(X(t) r) ~ p z ) (y(t ~ = E{z(t)~ (.~ b:xpy + = vXg( Z) ~ Solution 17.14 Forinirig the timeaverage according t. because the enserriblr means rorrcspond to the signal ilself.lly iincorrcla.tC} .tlierefore: ~~ ( d ( t ) p)’i = & ( t ).46) a i d y . { g ( t ) } ! or both expected expected values must be zero + (at. least onc of Ihe randoiii signal inust have zero rrieaiz).2pd(1) + p2 ~ = @ ( t ) 2/rd(t)  + /? = q i j .18) is linear. With p = cl(t).x(t) g ( l ) ) ’ } = E{:x2(t) + 2 r ( l ) y ( t ) y ” t ) } = + + = E(s2(t)} 2E{r(t)y(t)tE{&)} = E { 2 ( 2 ) } E { y Z ( t ) } } + + To sitt’isfy the condition: E{x(t)y(t)} = 0. F:{z(t)} /~~.p.8) Solution 17.riy c:liosen random processes are gcnern. Solution 17. /is = E ( : r ( t ) }yield: E{&)Y(~)) W4t)) E{y(tj) = c) Eitbcr E { z ( l ) )or b .
 . J L J ) : coiljugate symmetry.1) @tbL. 17. since ACF real Solution 17.l 1 9 ~ p[0] = E{z2[k]) = .?Llj o 1 ' I w 0?&LJ) = @t.p.3 B A a12 = (ynLr[~ E { 2 [ k ] } .?!)real (r even (311. 4 Solution 17. p T r[ K # 01 = pu. since pY._ .see Ex.18 $?. real eveii.] U ) : (.lL(Jwj p. Solutioris to the Exercises 551 Solution 17.'. ( + u(t) real. ~ 1.Appendix A.K]} k c.19 O+111+1 p 7 ..17 (with (9.1 = E { z [ k ] .(.12) arid Table 4.2 # 0 = 1 LL. 12 9 3 . c [. at.(.I 91 " Since the iiunihers of dice are uncwrelatcd with different time k .4 16: 16 loecaiise the outcome of a die does not correlate t o clifferelit = o'+ IL +12 ./'()d= LJ2 = ejl"uLJ 2 +1 + 2 +1 + even.1) and (DOLL .8.
2 ( r ) ) 1 7r . Q.552 Apprritlix A. Sohit ions to thP Exrrciscs 9 3 . r ( r = 51. 0 ( ~ ~ = ) . p m / 4= 16 + O[K] 16  Solution 17.4CF.20 ..recl 1 1 2 (z) (z) * rect p r .
Sohit ions to t. makes T t.Appendix A. Adding a constant.lie system noiiIiriwr (scr C'h. 1). brcaiisr we arc iiot coilsidering ail L T system.he Exercises 553 11) KO. .
8) and ( I 8 17)..23d _ 2 _ _ +_ ~_ + 2gw + 2) .( 4 = B"P. we ubtaiii: PZZ(.(T) + Solution 18.5 IJsing (18.6 a) H ( j w ) = ( p 2) (w'L + 4 . Solutions to the Exercises Solution 18.) = 1AI2P/4TL(4 IBI2Py..7).(4 With (18.554 Appendix A.21) arid (18. for uncorrelated signals ( P J Z ( 7 ) 2 and y 1iolds = A*it?rz(rj ( P 2 . .
lie Exrrriscs 5.lrise.Amendix A. Solutions to t. calculatioii of plrh(T ) can be arliiewd with thr shift theorem: .8 a ) Inverse Foiirirr transform with thr 5hift theorcm. 0 b) Lil<r.55 Solution 18.
556 Anurndix 4 . Soliitions to the Hxrrciscs Solution 18.9 ACF: .
any riurriber o f allpass stages will still not change l H ( j w ) I 2 .Apgeiidix A.10 557 b) No. Soliitioiis to the Exerciser S o l u t i o n 18. .
3 0. 0.7 0.6 Solution 18.8 0.7 1 and swapyiiig the sums yields: I I 201 0 '\ ' . ( j w )of the received signal y(t) and the crossspectrum between y ( t ) arld the tianwiittpd signal r ( t ) . . r .15 a) To determine the Wiener filter we need to h o w the powri density spPctilln1 Q g L .4 QI3l 0. . 'Yhe optinial solution to this exercise @ r u ( ~ A i ) b) Head thc linear distortion without nolqe from the block diagramm: G ( s )= 6I S + H ( j w ) = ___ us ~ 1 +a = +6 S G(3w) .2 0 0.9 1 0.1 0.5 i 0.558 Siihstitutirig d = / I A4nr)endix Solutions to the Exercises A.
Solutions t o the Exercises 559 .Appendix 12.
.put IS @it).17 The noiseflee signal at the out.).(~ C) ~ ( t t} + ) E{C(t + 7 )~ ( l )= 0 + pi. pg.efor.Jr) = E{y(l + rj ~ ( t )=)F { [ n ( t+ Z) + Y ( t + r ) ]~ ( t ) } = = E{. (18:78) p ~ : ~ (= ) ( z ) arid tl1er..() } T ZLI With ~p.W..e rh = W. ..ioris to the Exercises otlierwise Solution 18. Solut.’) = @‘1S.TI.(Ji.(.?LJ) (18.560 Apperrdix A.~(z) 6 ( r ) gilt analog = @.77).:...
. .(t) Kc{s} s 1 >0 + (1 Re{ s } > Re{u} t z (t) '1 Rr{s) > 0 Re{s} > Re{(]} COS (d.Appendix B Tables of Transformations Appendix B.s E 6 r .)( t ) t)E R{} > 0 e.1 Bilateral Laplace Transform Pairs KO(' .
2 Properties of the Bilateral Laplace Transform I Linearity z(t) .(t) Dclay :r (t .562 Appendix B.lables of Transformations Appendix B.U ) . Diffrrrntiation in the frequency domain Hr{ci} shifted 1 R e { u } to the rig21 iiot afleected t.4qjtj + Bz.hlultiplication by t'.L) X($.r(t) I I Uiffereiitiat.ion in the t iiiie doiiiain d lntcgration sX(s) ROC KOC{X} 2 I I" I 1 1 X(S) TC(7)dL H ROC 2 ROC{X} n j J : He(s j > 0) ROC' scaled by a fwtor of Scaling z(ut) 1 x In1 (i) .
Fourier Transform Pairs 563 Appendix B.3 Fourier Transform Pairs B‘(t) I J” si(at) 1  rec:t.I 27r6(Ld . 7r Id (E)  t sign (t ) I jmiign(iL?) 2 .3.Appendix B.U . ) ) .
cgr+ 1011 I' t ' = .iatioii in t.4 Properties of the Fourier Transform Linear1ty cJ iJ 7 x ( J iLI) Alodulatioii "Iulttplicat.he t i i i i p domain ts(/) Int.563 Appendix B. T.ion by t' Diffrrctritiat. a E IR\{O) Mu It ipl ica tion Dua lit? x(t) P(1) J"'k(t) Jx .hc frecluency clomain Dif€ereut.ioii in t.x ( p ) + 7rX(O)6(d) JLJ Scaling C:oIlvolut ion IX I4 )::( .tbles of 'lhiisformntions Appendix B.1.
5. Twosided zTransforni Pairs 565 Appendix B.5 Twosided zTransform Pairs .Appendix B .
566 Appendix R.Tables of Transforiiiatioiis Appendix B. scpsrate of consideratioii 2=0 Time invcrsion x[k] ROC= {Z I. Delav separate ul consideration z = 0 a i d s + rx.l~ROC{~)} multiply tlie limits of the ROC . Modulation Multiplication by k ROC{ r } .6 Property Properties of the xTransform J:[ k ] Lincarity ' ROC{ x } .
7.7 DiscreteTime Fourier Transform Pairs x[kj d[k] 1 COS Rok:  1 [ u ( R. +U(?)] R F) 2 + .Appmdix H.. DiscreteTime Fourier Transform Pairs 367 Appendix B.
tion Parsrval t.568 Appcndix E.8 Properties of the DiscreteTime Fourier Transform Property Lincari ty Dchy x[kh.ion Mult.] Mvclulat ioii Convolut. Tables of Transformat ioris Appendix B.iplica.heorrin one period! .
Sivan. K. Brrkelcy Carnbritlgr l m Prrss.Bibliography [9] W . 2nd edition.y crmd Lasncnr S ~e Ys .Heck.~Il. Signal Froce C!a.ri. S. 1998. Prriitic*eH. /. bhgle. Englewootl Cliffs. Karii. [Ill H. lnfrod?iction to Signals and Systems. 'IliCB/n. Funda.lcCr. [12] B.e. Bost. 1909.tals of Sbgn. USA.P. 199 1.Linclrier.Kwakrrnaalc arid R.Z. 1997.lriood Cliffs. Lattii.dl.as~~Hill.trl. [ 131 D. . Signals and Syslems A9ade Rzdiculously Simple. 19%.uls and Systems Usa. [10] Z. Eiarri~iiarid B. Prentlct'H. Modern Signalv uml Syaterns. ZiZiPress. MA.oii.rri.rmich.ng thr Wf:hnnd Matlab. Cainhndgc.
Wuusch. 2. Netzwelle. cmd Systems. RIcCraw IIill. . Teil I. StgrLuJe . Sigrsuls.G. 1990. Berlin. T MA. NewYork: 1983. VCH Verlagsgesellschnft .zmeal. [17! R. ~ [18] 1I. des SpringerVerlag Uerliri. Basel. rofech. [“O] W. 1986. 1973. Oldcnhourg Vcrhg. Bcrlin. Berlin. TJSA. . 7 edition.Mc~C. I96 7. 1994. Oppenheim and A. Reid. Willsly.d S’ysteme 1. The M T Press.W Schiifjler. Unbehauen. Grundlnge. 1997. Florida. Oltlmbourg Verlag. CRC Press. 124) G. [15] A.i7.nd T.un. L. Geschzchte d e r Systerntheor. Nctzwer 3 edition.rr der El edition.w.nik. Signde irnd S~yste~rne SpringcrVerlag.. Bcrlin. Badii.[14] F Oberhettinger aiid L. h‘liilxchCi1. SpringerVerlag. Sauer a. Signale und Systcme. [19] H. Cambridge. Bahill. 2 cditioii. S y s t e m Pundanzentnls. Szah6. SpringerVerlag. [22] R.Systems Tlieory. Springcr Verlag. [16] J. 1991. Lznear.W. 1U92. 4 [23] R.e.Ci~cutts. Unbeliaueri. Schiijjlcr. 1989.eo. Tuhles of Laplace Transforms.cllzsche I€~. Mutherri. Systenith. 1983. Siehert~. Bocn Itaton. Miinchen.l$smittel Ingenwurs. [2l] F Szidamvsky a n d A.T.
318. Y 9 O shift properly. 405 bandlimit ed signal. 254 splane. 308 modulation property. 405 along the process. 286 analogtodigital converter. 179 P . 330 iriverse.2riancel 420."transform. 405 linear. 276 bandwid t 21.urc correction filtcr.v iritrgral. 271 baldpass rignals complex.r frequency. 282 funclion. 48: 325.l. 241. 231 bascband. 92 caural. 112 causality.odigital c:onvert. 407. 87 spectral. 261 aliasing. 275 rcal. 271 basis funclioiis. 2 19 angula. 21 Cauch. 388 .ioii theoreill. 298 angular sampling frequency.405 time. 410. 281 autocorrelation. 228. 19.oc:o\l.plane.ransform convolution property. 443 aiit. see autocorrelation function. 442 quadratic! 407 stat. 269 BIB0stability. 424 sequences.c k u i t . 308 nmltiplicat. 322 convolution theorem: 353 firiite seqiience. 326 F t. 315.he process.er. 325 npert. bee analogt. 330 infinit. 28 I .Index DTIcircuit. 367 chnracteiistic seqwnrc. 410 ADC. 397 Qfaclor. 331 autocorrelation f i r d o n . 3 3 cwiionical form. 425 average across t.istic:a. 272 timedoniaiii. 26 I analysis c:omplcs. 383 biliriear t r arisform. 31b coriip1ex ACF.c sequence. 388 . 317. 317 shift theorem. 308 rectangular. 328 of an exporieritial sequence. 392 block diagram. 319. 325. 336 Bode tliagrwm. 316. 271 basrhwid rpcctrimi. 344 theorems.
/I03 ol cornplex signals. 191 delay circmt. 302 discrete unit st. 5 DFT. 431 croscovariaiice. 344 discrete Foiirier traiisforrri. 426 crossspectruin.rrris. 309. 126 specific: solution. 268 tliroreni.iat. 24 discrcte convolution. 80 dilTeereiit. 37 controllable system. 301) dampiiig. 298 tlisc:retje uriit iiiipulse. convolution tlieorem. 5 D i r x delta function.udc. 51 convolution.or. 295 tiecoiivcJli~tictii. 182 cyclic. I26 diffc. 35 controller. 248:251 cyclic convolution. 48 freyiirncy plme. 301 discretetime signal.r c. 36. 286 discrek step fitnction. 352.46 frequency parameter. 296 diniensionalit. 298 haiidpa.Jl~fOrll1..ombination. 297. 445 r11toff frequency. 352. forin II. 217 correlation function. 303 discretetirrir Fourier lraiisforin. 309 discrete. 50 cxponcmt>ial signal. 49. 425 crosspower derisit.t~ioii theorem. 20. 110 c:oinplrx a. 171. 18 Ixwtial. 48 complex pole pair. 20.l signal. 346 direct. 17 ordinary. 79. 297. 88 complrx freyucncy plane. 169.l fiiiiction. 268. 46 hX'ClUeIic\7. 18 tIiffcrrntia. 197 coniplex mslysis.e F'onrier ti"a. 249. 417 correlation functions. 351 b y inspection.e delay circuit. 87 main principle of. 158 direct. 339 distribution. 5 duality. 421. 426 cimscorrealtion function. 177. 158 Dirac irripiilse.rrntia. 350.ion. 266. 34 1 differential equation. 350. 165 dctcrmiaism. 162 linea.ion. 358. 423 critical sampling. 356 diacrct.y spwtriini. 1% liomogcnous solut.ep function. 3.y. 171. 175. 268. 160 drrivatmive. 215 . 423 seqnences. 153 domain. 178 controllability. 346 direct Forin 111.ical solution. 396 romrcrgcnce'. 318 niialyt. 273 crosscorrclat.351 cdculatioii rules. 161 derivation. 2 x 3 . 33 1 riiiinrrical solution. 176 lr1dt2 d r k a impulse. 175 27. 423. 170. see discret.nipliliide spectrum. 356 pcriodic.l equations. 286 dilferrnce rqiiation: 340. 227. 77. 210 contiiiuity. 254 cwijugatc symmetry. 261. digita. 5 control. form I. 2!)5 discretetime syst. 309. 162. 36 controlldde. 18 with constant rorfficicnts.14. 169.ss signals: 275 exponeiit1a.572 aiiiplit.
riiiii. :369 .: 42 I cxpect. 160. 1 . 5153 cigensequence.ial sequriice.ioa. 4 1 341 Fx transform. 39S filter autocorrelahri fiinct. 234 Hilhcrt transform. 301 fccedback. 341 esiergy of a time signal. 298 undamped. 254 frequency parameter.irnill. 286 tfiscretet. 414 error power. 5 . 226 ensemble.tched. 350. 266 Fouriu series. 444 rna. 264 shift. 209 holomorphic.2 . 392 Ic:ircuit8.inie. 1110 exponential func t.rity. 301.iori 52 cxponeiitial ordcr. 200 periodic: signals.ion.171 tier 573 differentation t hcorein. 69 rveii. 72 exponent. 453 fiker ACF.in: 45. 209 regu1a.l. 309 frequency response. 310 expected value. 407 secoridordcr. 215 rnodiilat.e. 46 freciuencytlorii.iablc. 453 Wicsier. 174. 241. 221 rnultiI)lic:.~ G9 esgenfisnctmii. 2 17 frcqucncy resonnnt. complcx. 301 duality. 48. 69 odd. 322 convolution. 217 Gauss impulse. 319 unilateral.ed values firstorder. 188 opt. 196. coinplcx! 302 external part.tt.ion. 356 snipedance. 353 estirnat#ionerror. 241. 61. 197 Fuiiriw transform. . 333 smoot.i. 200. 69. 351. 221 siiiiila.ic. 167. 452. 304 exponential srqisences. 269 iiitjegratlsosi tlieoreiis. 444 FIRsysteni. 38 freqisriicy resolutsosi. 61. 301. 88 complex tliffcrent. 441 joint. 396 frequency. 405 secondorder joint. 322 pairs. 372 Hurwitz deterinsnants. 128. 209 even srqucnce.179 IIRsystem.219 of a sequence. 356 Fourier cocfficierits. 3% polynornial. I3 2 inverse discrvtctime! 30 1 lisrearity. 224 inverse?. 85 fun ctioii analyt. 197. 404 rnsemblc siieaii! 405 ergodic. 322 F?: tjransforniirivrrsc. 191. 303. 453 even function. 215. 261 Fourier spect. 422 ergodic random processes. 317. cosiiplex.liing. 224 discret. 5 3 hipulse respoiise.
116 series coupling. 115 LTIsystriiis conibinat ion. 404 normalisation. 37 observable svstem. 177 uinila. 142 integralile. 72. 76. 70. 310 operational amplifier. 34 system. 76 practical calculation of tlic iiiverse. 36 obsc. 126 firsborder. I42 init. 129.rv. 395 Laplace integral. 64 Laplace Transform inverse bilateral. 61. 79 Laureiit expansion. 97 lap lac^ t. 198. 102 modulation theorem. 54 Nyqiiist frequency. 87. 3 3 transformation. 348 cla. 72 integration theorem. 384 iiitegrat.ial value. 5 mean square. 188 matrix Frobeniiis. 128. 388 coriibination. 61.teral! 61. 77. 175 existence ol. 327 noise powei . 135 with smusoidal sigiial. 220 odd function. 73. 327 iridtiplication property. 96 inverse with coinplcting the s y a .ransforni. 6366. 62. 132 initial state. 144. 197 matched filter. 324 bila. r c . 21Y multiplicat. 28. 127. 266 linearity. 75 shift theorcm. 7 lowpass f i h r ideal.445 measiiremrnt of the iriipulae respoiise: 451 of the transfcr function.451 modal niatrix.heoreim. 322. 125. 369 causal stable. 69 line spectrum. 98 inverse unilateral. 261 indefinite iiltegral. 30 completely. 10. 30.er. 34 I intjerpolation filt. 453 . 175 irivrrse. 340 feedback. 221. 19. 33 niodulation. 76. 77. 195. 68 iiiilial condition. 232 rnagnitudc spectrum. 129.ion tlieorerri: 77. 63. 101 properties.l state: 143 Inversion. 79 differentiation t. 117 parallel coupling.tblr. 386 Index LTIsysteiri. 273 obscrvabilitv.teral. 271 intia. 129 iriilinl c:onditioii problem. 33 meari. 19 natural. 33 modal. 28 optimal filter. . 179 magnitude lreyuency response. 177.ssica1solution. 20Y odd sequence.574 irnpulse train. 264.hcorem. 79 integrator. 222 modiilation t. 174 iiiteriial p r t .ion theorem. 115 discrete. 62. 341. 35 obser vat ioii window of.
5 . 104 raiidoni sequences. 407. 272 sampling theorem. 285 idcal. 385 of the zt. 2. 179 yuanlisat.ion. 1 amplitide. 69. 201 202. 275 contimious. 254 rms. 1 a~niplitudccontinuous.erniinist.19. 62. 447 PT1circuit. 261. 241.ioii. 3 1 5 . 320 residue (.h a. 1. 197 phase spectrum2 197 PIcircuit. 261 i lime. 404 sampleandhold. 4/15 power density spectrum. 452 rectangle function.5 ariiplit. 445 ROC'. 273 Pcircuit. 262 shift theorern. 2. 437 real bandpass signal.1 polezero diagram.uc~ediscrctc. 376 frecluencydorriaiii. 376 critic:al.ed. 110 pole pair. 98 for multiple poles.5 order of a systein. 179 pole. 31 oversampling. 80 . signal. 1 ckt. 282 of a.mp1t ude.Irrde2: 57. 297 sequcncc even. 170. 264 phase. 376 bantllimit. 67. 325 sclective property. 155. 253 resonant frequencl . 235 partial fraction expansiori. 438 miilt'iplicatiorl wit.ion. 4.ransform. 296 discontiiiuous. 282 snmpling. 397 sariiple €unct. 31 periodic: convolution. 269 real ~ o r l d 279 . 5 analyticaS. 225. 431 rnndoiii signal addition. 91 ~ resonaiice cwrve. 296 a. 269. sampling frequency. 261. 5 analogue. 5 pcriodic. 204 rectangle inzpulse. 70. 203 signal. 5: 403 digital. 264 piecewise continuous.symbol. 269 sampling raw.time. 3 I0 odd. 64 66. 64. 111 output equation. 271 causal. 159. 202 rcgion u l convergericc. complex. 22 1 . 3 3 3 power. 1. 270. 34 Parseval relationship. 5 discretetime. 2R5 mult. 62 root locus. 261 random process. 261. 265 pcriodic signal. 273.idirnensional~ 3 4. 4. 276 realisat. 261. 404 rcconstructiou. 110. 25. 332. 4 continuous. factor. 310 Parseval's thcorml. 179 parallel form. 76. 327 sifunction. 353.ic.heorem. 103 path direct.12. 370 complcx bandpass. 310 sha. onedimensional. 295.
343 rat1011a.iaiit. 64. 5 callsal.aliun. 339 matrix. 174 suporposition priiicipk. 197 Fouricr. 21 7. 29 eqllat~lon. 271 coiiiplex amplitude.nce.r. 3 3 tell1 stat.)anrlwitlt. 8. 447 st. 13 rnemoryless. 118 trniisfoim . 37 Indel. I97 of a sequence. 385 state.trix. 315 ~ lliirwitz crit. 383 BIBO. 107 ten1 ina. 110 teni idorrtification. 50.7 : 9 t. 413 st at 1stica. 107 systtin fiinction: 105. 197 power densitmy. 169. 174 derivat. 165 step response.ystcm. 371 baseband. 405 t. 233 timedoiiiaiii aliasing.ion. 286 kderance srhcme. 13 system analysis. 374 phase. 12 time vana.ria. 295 timeaverage 43 1 sccondorder joiiit. 138 represerit. 395 nonrcciirsivc. 164.e. 65. 339 equivalent.l analysis. 405 Step fUIlCt~ioK1. 276 recoiistrucliori of. 383 c:lnssificnt. 231. 34 melnory. 347 stationary. 112.eria. 13 minimal phase. 7. 339 discrctetime. 53 shiftiiivariaiit. FIR.z.l fract1o11.63. 7. 339 syniiiretry coiljugat c. 452 stoc:li.iou. 219. 327 tiinc series. 4. 395 linea.ion. 422 tinicI.. 308. 356 nonlinear. 356 funct. 327 singularity. 9. 233 transfer fiinction. 340 lranslationiii\. 412. 367 . 396 time invariant. 367. 125 time average.13 inverse. 107. 1 3 tlifferent. 219 spectrum. ( 9 5 spectra.imc iiiva. 69. 340 stabilising. 2 10 s. 13. 31. 301 oncsided. 220.liproduct. 105.iit.irricinvnriant.ar. 356 iriversiori. 89: 90 essential. 422 weak. 404 similarity throreiri. 391 stahilily crikria . 441 joint. 12 t. 29 iiiatrix733 variable. 3. 13 discrete. 50 IIR. 356 respoiise.576 real bandpass.ial eqimtiori. 7 recursive. 108.l averagc.iiiie reversal.abilising: 396 stability. 197 magnitude.tstic.
aplnce. 139 uricerlaiiity rclatiori.lic Laplace tmnsform: 83 1. 220 f zero. 73 iiiiit. 453 window o observat./rider 577 . 19 I 452. 105 transform pairs of t. 1 variaiiw. 407 whit. 2 dependent. 110 . 273 uniqueness. 61. 1 intlcpeiiclerit.F. 5.raiisLoririetion parallcl forin. 236 uncorrclated.ion.c noise. 61. 297 discrcte. 36 tmnsformalioii ma. 158. 62. 372 I. 196 Hilherl. impulse. 388 unil. 302 variable coiitinuoits.trix: 33.+ . 392 Foiirier. circlc. 301 bilinear. 430 Wiener filter. 318 uridersairipling. 429 baiitllimited.
This action might not be possible to undo. Are you sure you want to continue?