This action might not be possible to undo. Are you sure you want to continue?
Basic Deﬁnitions and The Spectral Estimation Problem
Lecture 1
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L1–1
Informal Deﬁnition of Spectral Estimation
Given: A ﬁnite record of a signal. Determine: The distribution of signal power over frequency.
signal t ω t=1, 2, ... ω ω+∆ω π spectral density
! = (angular) frequency in radians/(sampling interval) f = !=2 = frequency in cycles/(sampling interval)
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L1–2
Applications Temporal Spectral Analysis Vibration monitoring and fault detection Hidden periodicity ﬁnding Speech processing and audio devices Medical diagnosis Seismology and ground movement study Control systems design Radar, Sonar
Spatial Spectral Analysis Source location using sensor arrays
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997 Slide L1–3
Moses.1 = t= If: discretetime deterministic data sequence X 1 t=. Prentice Hall.i!t exists and is called the DiscreteTime Fourier Transform (DTFT) Lecture notes to accompany Introduction to Spectral Analysis by P.1 y(t)e.1 jy(t)j2 < 1 X Then: Y (! ) = 1 t=. 1997 Slide L1–4 .Deterministic Signals fy(t)g1 . Stoica and R.
S (!)d! t=.1 X (k)e. Moses.1 X Z 1 where 4 S (!) = jY (!)j2 = Energy Spectral Density We can write S (!) = where X 1 k=. k) Slide L1–5 Lecture notes to accompany Introduction to Spectral Analysis by P.1 y(t)y (t . 1997 .i!k (k) = 1 t=. Stoica and R. Prentice Hall.Energy Spectral Density Parseval's Equality: jy(t)j2 = 21 .
1997 Slide L1–6 .1 n jy(t)j2 = 1 o But: E jy(t)j2 < 1 E f g = Expectation over the ensemble of realizations E jy(t)j2 n o = Average power in y (t) PSD = (Average) power spectral density Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R.Random Signals Random Signal random signal probabilistic statements about future variations t current observation time Here: X 1 t=. Moses. Prentice Hall.
Prentice Hall. Moses. (!)ei!k d! n o (Inverse DTFT) Interpretation: so r(0) = E jy(t)j2 = 21 .1 r(k)e. (!)d! Z (!)d! = inﬁnitesimal signal power in the band ! d! 2 Lecture notes to accompany Introduction to Spectral Analysis by P.k) r(0) jr(k)j Z Note that r(k) = 21 .i!k (k) is the autocovariance sequence (ACS) r(k) = E fy(t)y (t . k)g r(k) = r (. Stoica and R.First Deﬁnition of PSD (! ) = where r X 1 k=. 1997 Slide L1–7 .
i!t (!) = Nlim E N !1 t=1 X > : 8 > < 2 9 > = > Note that 1 jY (!)j2 (!) = Nlim E N N !1 YN (!) = X where N is the ﬁnite DTFT of fy(t)g. t=1 y(t)e. Moses. 1997 Slide L1–8 .i!t Lecture notes to accompany Introduction to Spectral Analysis by P.Second Deﬁnition of PSD 1 N y(t)e. Prentice Hall. Stoica and R.
P2: ] () f 2 . Thus.! ) Otherwise: (! ) 6= (.! ) P3: If y Lecture notes to accompany Introduction to Spectral Analysis by P. Then: (! ) = (. Stoica and R. Moses.1=2 1=2] (!) 0 (t) is real. 1997 Slide L1–9 . we can restrict attention to !2 .Properties of the PSD P1: (!) = (! + 2 ) for all !. Prentice Hall.
Moses. 1997 Slide L1–10 .1 H (q) = X 1 k=0 hk q. k) hk e.Transfer of PSD Through Linear Systems System Function: where q . 1) H (q ) y (t) 2 y (! ) = jH (! )j e( e(t) e (! ) Then !) y(t) = X 1 k=0 hk e(t .1y(t) = y(t . Prentice Hall.i!k H (! ) = X 1 y (!) = jH (!)j2 e(!) k=0 Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R.k = unit delay operator: q.
]g Find an estimate of Two Main Approaches : Nonparametric: – Derived from the PSD deﬁnitions. Moses. Prentice Hall. 1997 Slide L1–11 .The Spectral Estimation Problem The Problem: From a sample fy(1) : : : y(N )g (!): f ^(!) ! 2 . Stoica and R. Parametric: – Assumes a parameterized functional form of the PSD Lecture notes to accompany Introduction to Spectral Analysis by P.
Prentice Hall. Stoica and R. Moses. 1997 Slide L2–1 .Periodogram and Correlogram Methods Lecture 2 Lecture notes to accompany Introduction to Spectral Analysis by P.
Moses.i!t (!) = Nlim E N !1 t=1 X > : 8 > < 2 9 > = > Given : Drop “ fy(t)gN t=1 lim ” and “E N !1 f g” to get X ^p(!) = 1 y(t)e.Periodogram Recall 2nd deﬁnition of (! ): 1 N y(t)e.i!t N t=1 Natural estimator N 2 Used by Schuster ( 1900) to determine “hidden periodicities” (hence the name). 1997 Slide L2–2 . Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Prentice Hall.
1997 Slide L2–3 .(N .1) Lecture notes to accompany Introduction to Spectral Analysis by P.1 X r(k)e. Moses.1 N .i!k ^c(!) = k=. Stoica and R.Correlogram Recall 1st deﬁnition of (! ): (! ) = P X 1 Truncate the “ ” and replace “r k=. Prentice Hall.i!k (k)” by “^(k)”: r r ^(k)e.
1997 Slide L2–4 . k) k 0 r(k) = N ^ t=k+1 X For both estimators: r ^(k) = ^ (.k) k < 0 r Lecture notes to accompany Introduction to Spectral Analysis by P. k) k 0 ^ r(k) = N . Prentice Hall. Stoica and R.Covariance Estimators (or Sample Covariances) Standard unbiased estimate: 1 N y(t)y (t . Moses. k t=k+1 X Standard biased estimate: 1 N y(t)y (t .
Slide L2–5 Lecture notes to accompany Introduction to Spectral Analysis by P.i!t N t=1 X 2 = N . Prentice Hall. Moses. Stoica and R.i!k ^p(!) = ^c(!) Consequence: Both p ! and ^( ) ^c(!) can be analyzed simultaneously.1) r ^(k)e.1 X = ^c(!) k=.(N .Relationship Between ^p (! ) and ^c(! ) If: the biased ACS estimator r Then: ^(k) is used in ^c(!). N ^p(!) = 1 y(t)e. 1997 .
Stoica and R. 1997 Slide L2–6 . Thus. Intuitive explanation: r(k) . ^p(!) and ^c(!) have poor performance. r(k) may be large for large jkj ^ Even if the errors r k r k Nj. the PSD error is large.1 are small. Prentice Hall. ( )] Lecture notes to accompany Introduction to Spectral Analysis by P. even for large N . Moses.Statistical Performance of ^p(! ) and ^c(! ) Summary: Both are asymptotically (for large N ) unbiased: E ^p(!) ! (!) as N ! 1 n o Both have “large” variance. p! f^( ) . ( )g ^ ( ) . jk =0 there are “so many” that when summed in ! .
Stoica and R. n o 0 k 1 . 1 jkj N Bartlett. window E ^p(!) = 21 . ( )WB (! .(N . N k=.i!k = 1.1 n o n o X ! X N . Prentice Hall.1 X wB (k) = 8 < : = Thus. jNj jkj N .i!k k=.Bias Analysis of the Periodogram E ^p(!) = E ^c(!) = E f^(k)g e.1 jkj r(k)e.1) N . ) d Z Ideally: WB (!) = Dirac impulse (!). 1997 .i!k r k=. Slide L2–7 Lecture notes to accompany Introduction to Spectral Analysis by P.(N .1) 1 = wB (k)r(k)e. Moses. or triangular.
Prentice Hall.Bartlett Window WB (! ) 1 sin(!N=2) 2 WB (!) = N sin(!=2) " # 0 WB (!)=WB (0). Moses. 1997 . Stoica and R. Slide L2–8 Lecture notes to accompany Introduction to Spectral Analysis by P. (!) may differ quite a bit from (!). WB 1=N . for N = 25 −10 −20 dB −30 −40 −50 −60 −3 −2 −1 0 1 ANGULAR FREQUENCY 2 3 Main lobe 3dB width For “small” N .
Smearing and Leakage Main Lobe Width: smearing or smoothing Details in resolvable. Stoica and R. Prentice Hall. φ(ω) smearing ω ω (!) separated in f by less than 1=N are not ^ φ(ω) <1/Ν Thus: Periodogram resolution limit = 1=N . Moses. 1997 Slide L2–9 . ^ φ(ω)≅W (ω) B Sidelobe Level: leakage φ(ω)≅δ(ω) leakage ω ω Lecture notes to accompany Introduction to Spectral Analysis by P.
!1 ^( ) ( )! ( ) Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall. Stoica and R. Moses. severe bias As N . 1997 Slide L2–10 . so ! is asymptotically unbiased.Periodogram Bias Properties Summary of Periodogram Bias Properties: For “small” N . WB ! !.
Stoica and R. dev = φ(ω) too  Resolvability properties depend on both bias and variance. (!2) 2(!1) !1 = !2 = 0 !1 6= !2 nh ih ( io Inconsistent estimate Erratic behavior ^ φ(ω) ^ φ(ω) asymptotic mean = φ(ω) ω + 1 st. (!1) ^p(!2) . Moses.Periodogram Variance As N !1 E ^p(!1) . Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 Slide L2–11 . Prentice Hall.
Stoica and R. Prentice Hall.i!t 2 k and W = e. 1997 Slide L2–12 . Moses.Discrete Fourier Transform (DFT) Finite DTFT: YN (!) = X N t=1 y(t)e.i 2 . 1 Direct computation of O N 2 ﬂops ( ) . N Let ! = N Then YN ( 2 k) is the Discrete Fourier Transform (DFT): N X Y (k) = N t=1 y(t)W tk k = 0 ::: N .1 fY (k)gN=0 from fy(t)gN : t=1 k Lecture notes to accompany Introduction to Spectral Analysis by P.
2 = 2p: Y (2p) = For k X ~ N = 1 3 5 : : : N .i2 =N . y(t + N )]W tgW tp Lecture notes to accompany Introduction to Spectral Analysis by P.1 for odd k Assume: X X X Let N For k ~ ~ = N=2 and W = W 2 = e. Moses. Prentice Hall. ~ 4 = 0 2 4 : : : N . Slide L2–13 t=1 ~ ~ f y(t) . 1997 . Stoica and R.Radix–2 Fast Fourier Transform (FFT) N = 2m N=2 N Y (k) = y(t)W tk + y(t)W tk t=1 t=N=2+1 N=2 2 = y(t) + y(t + N=2)W Nk ]W tk t=1 Nk +1 for even k with W 2 = . 1 = 2p + 1: X t=1 ~ ~ y(t) + y(t + N )]W tp Y (2p + 1) = Each is a N ~ N ~ = N=2point DFT computation.
Prentice Hall. Stoica and R.FFT Computation Count Let ck Then = number of ﬂops for N = 2k point FFT. 1997 Slide L2–14 . Lecture notes to accompany Introduction to Spectral Analysis by P.1 k2k ) ck = 2 ck = 1 N log2 N 2 Thus. 2k + 2c ck = 2 k. Moses.
N=8 ω 2 k N . 1997 Slide L2–15 .1 N k=0 sampled. Moses. Prentice Hall.Zero Padding Append the given data by zeros prior to computing DFT (or FFT): fy(1) : : : y(N ) 0 : : : 0g  N {z } Goals: Apply a radix2 FFT (so N Finer sampling of = power of 2) ^(!): 2 k N . Stoica and R.1 ! N k=0 ^ φ(ω) continuous curve Lecture notes to accompany Introduction to Spectral Analysis by P.
1997 Slide L3–1 . Stoica and R. Prentice Hall.Improved PeriodogramBased Methods Lecture 3 Lecture notes to accompany Introduction to Spectral Analysis by P. Moses.
Stoica and R.i!k r fw(k)g = Lag Window w(k) 1 M M k Lecture notes to accompany Introduction to Spectral Analysis by P.(M . Prentice Hall. 1997 Slide L3–2 . ^( ) jj ^BT (!) = M .1) w(k)^(k)e. Moses. with small weight applied to covariances r k with “large” k .BlackmanTukey Method Basic Idea: Weighted correlogram.1 X k=.
)d 2 . Stoica and R. Z W (!) = = Conclusion: Effect: DTFT fw(k)g Spectral Window ^BT (!) = “locally” smoothed periodogram Variance decreases substantially Bias increases slightly By proper choice of M : MSE = var + bias2 ! 0 as N ! 1 Slide L3–3 Lecture notes to accompany Introduction to Spectral Analysis by P. Moses. con't ^BT (!) = 1 ^p( )W (! . Prentice Hall.BlackmanTukey Method. 1997 .
Stoica and R.Window Design Considerations Nonnegativeness: ^BT (!) = 1 ^p( ) W (! .(M (0) w Z w (k ) = equiv time width 1 2 . 1997 Slide L3–4 . Prentice Hall.1 X . )d 2 . 0 Z  {z } If W (!) 0 (. Moses.1) Ne = k=. W (!)d! = equiv bandwidth e= W (0) Ne e = 1 Lecture notes to accompany Introduction to Spectral Analysis by P. w(k) is a psd sequence) ^BT (!) 0 (which is desirable) Then: TimeBandwidth Product M .
e) is essentially ) Choose window shape to compromise between smearing (main lobe width) and leakage (sidelobe level). ) Choose M as a tradeoff between variance and bias. bias decreases and variance increases. 1997 Slide L3–5 . As M increases. once M is given. Ne (and hence ﬁxed. con't e = 1=Ne = 0(1=M ) is the BT resolution threshold. The energy in the main lobe and in the sidelobes cannot be reduced simultaneously. Stoica and R. Prentice Hall. Once M is given. Moses.Window Design. Lecture notes to accompany Introduction to Spectral Analysis by P.
Prentice Hall. Stoica and R. Moses. 1997 Slide L3–6 . M 0 −10 = 25 −20 dB −30 −40 −50 −60 −3 −2 −1 0 1 ANGULAR FREQUENCY 2 3 Lecture notes to accompany Introduction to Spectral Analysis by P. M 0 = 25 −10 −20 dB −30 −40 −50 −60 −3 −2 −1 0 1 ANGULAR FREQUENCY 2 3 Rectangular Window.Window Examples Triangular Window.
Bartlett Method Basic Idea:
1 Μ 2Μ ... Ν
^1(ω) ^2(ω) φ φ
... average
^L(ω) φ
^B(ω) φ
Mathematically:
yj (t) = y((j ; 1)M + t) t = 1 : : : M = the j th subsequence 4 (j = 1 : : : L = N=M ])
^j (!) = 1 yj (t)e;i!t M t=1
X
M
2
1 L ^ (!) ^B (!) = L j =1 j
X
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L3–7
Comparison of Bartlett and BlackmanTukey Estimates
1 L ^B (!) = L j =1
X
8 > < > :
M ;1 X
=
'
Thus:
k=;(M ;1) M ;1 X k=;(M ;1)
k=;(M ;1) 8 M ;1 < 1 X L X
:
r ^j (k)e;i!k
9 =
9 > = >
L j =1
r ^j (k) e;i!k
r ^(k)e;i!k
^B (!) ' ^BT (!)
Since B ! implicitly uses method has
with a rectangular lag window wR k
()
^( )
fwR(k)g, the Bartlett
High resolution (little smearing) Large leakage and relatively large variance
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997 Slide L3–8
Welch Method Similar to Bartlett method, but allow overlap of subsequences (gives more subsequences, and thus “better” averaging) use data window for each periodogram; gives mainlobesidelobe tradeoff capability
1
2 . . . subseq subseq #1 #2 subseq #S
N
Let S # of subsequences of length M . (Overlapping means S > N=M “better averaging”.)
=
])
Additional ﬂexibility: The data in each subsequence are weighted by a temporal window Welch is approximately equal to nonrectangular lag window.
^BT (!) with a
Slide L3–9
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
for N 1. k +J 1 ^D(!k ) = ^p (!j ) 2J + 1 j=k. Stoica and R. Prentice Hall.J X Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 Slide L3–10 .1 ! = N j =0 Idea: “Local averaging” of (2J + 1) samples in the j (2 + 1) frequency domain should reduce the variance by about J . f ^p(!j )g are (nearly) uncorrelated random variables for 2 j N .Daniell Method By a previous result. Moses.
1997 Slide L3–11 . Hence: ^D (! ) ' ^BT (! ) with a rectangular spectral Let Z window. Stoica and R. Lecture notes to accompany Introduction to Spectral Analysis by P.Daniell Method. ^D(!) ' 1 ^p(!)d! 2 . Moses. Prentice Hall. for N 1. con't As J increases: Bias increases (more smoothing) Variance decreases (more averaging) = 2J=N . Then.
1997 . ^p(!) by a suitably selected ^(k) using a lag – Implemented by truncating and weighting r window in c ! ^( ) Bartlett. Moses. Prentice Hall. Stoica and R. Lecture notes to accompany Introduction to Spectral Analysis by P. Daniell Periodogram – Approximate interpretation: spectral window. Welch periodograms – Approximate interpretation: BT ! with a suitable lag window (rectangular for Bartlett.Summary of Periodogram Methods Unwindowed periodogram – reasonable bias – unacceptable variance Modiﬁed periodograms – Attempt to reduce the variance at the expense of (slightly) increasing the bias. ^BT (!) with a rectangular Slide L3–12 – Implemented by local averaging of periodogram values. BT periodogram – Local smoothing/averaging of spectral window. ^ () – Implemented by averaging subsample periodograms. more general for Welch).
1997 Slide L4–1 . Prentice Hall. Moses.Parametric Methods for Rational Spectra Lecture 4 Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R.
Moses. 1997 Slide L4–2 ( ) . Stoica and R. P By Weierstrass theorem.i! .θ) φ(ω) possibly revise assumption on φ(ω) Rational Spectra jkj m k e.i!k (!) = P jkj n k e. Prentice Hall. however: choice of m and n is not simple some PSDs are not continuous Lecture notes to accompany Introduction to Spectral Analysis by P.Basic Idea of Parametric Spectral Estimation Observed Data Assumed functional form of φ(ω. provided m and n are chosen sufﬁciently large. ! can approximate arbitrarily well any continuous PSD.θ) ^ θ Estimate PSD ^ ^ =φ(ω. Note.θ) Estimate parameters in φ(ω.i!k (!) is a rational function in e.
.AR.1 + + bmz. and ARMA Models By Spectral Factorization theorem. 1997 . A(! ) = A(z )jz =ei! Signal Modeling Interpretation: e(t) e (! ) = 2  B (q) A(q) white noise B (!) 2 2 A(!) ltered white noise y(t)y (! ) = ARMA: AR: MA: A(q)y(t) = B (q)e(t) A(q)y(t) = e(t) y(t) = B (q)e(t) Slide L4–3 Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall.n B (z) = 1 + b1z.g.1 + + anz. Moses. MA. e.m and. a rational factored as (!) can be B (!) 2 2 (!) = A(!) A(z) = 1 + a1z. Stoica and R.
k = j =0 = 0 for k > m B (q) = 1 h q. 1997 Slide L4–4 .ARMA Covariance Structure ARMA signal model: y(t)+ X n i=1 n aiy(t .k (h = 1) where H (q ) = A(q ) 0 k k=0 X X i=1 air(k . i) = X m j =0 bj e(t . j )y (t . Stoica and R. j ) (b0 = 1) Multiply by y (t . i) = m j =0 bj E fe(t . k)g Lecture notes to accompany Introduction to Spectral Analysis by P. k) and take E f g to give: X r (k ) + X 2 m bj hj . Prentice Hall. Moses.
.AR Signals: YuleWalker Equations AR: m = 0.1) : : : r(. Stoica and R. r(1) r(0) . Lecture notes to accompany Introduction to Spectral Analysis by P.n) . Prentice Hall. r(n) : : : r(0) 1 = 2 R 0 2 6 6 6 4 7 7 7 5 " # " 32 6 6 6 4 3 7 7 7 5 # These are the Yule–Walker (YW) Equations. . . 1997 Slide L4–5 . . an 0 3 7 7 7 5 2 6 6 6 4 Writing covariance equation in matrix form for k : : : n: =1 r(0) r(. Moses. . . .1) . r (. 2 1 a1 = 0 .
. . 1997 Slide L4–6 .1) : : : ^(. .1 = 0 a . r . r ^(1) r(0) ^ ^. r ^(n) : : : r(0) ^ ^n a 0 32 7 7 7 5 6 6 6 4 3 7 7 7 5 2 6 6 6 4 3 7 7 7 5 Then the PSD estimate is ^2 ^(!) = ^ 2 jA(!)j Lecture notes to accompany Introduction to Spectral Analysis by P. Moses. ^(. .AR Spectral Estimation: YW Method YuleWalker Method: Replace r 2 6 6 6 4 (k) by r(k) and solve for f^ig and ^2: ^ a r ^(0) ^(.1) . Stoica and R. .n) 1 r r ^2 . Prentice Hall. .
(Y Y ). 1997 Slide L4–7 .AR Spectral Estimation: LS Method Least Squares Method: e(t) = y(t) + X n 4 = y(t) + y(t) ^ where '(t) = y (t . ^ = . . . Stoica and R. . 2) Lecture notes to accompany Introduction to Spectral Analysis by P.1(Y y) where 3 7 5 2 6 t=n+1 je(t)j2 y(1) y(2) y(N . Prentice Hall. y(N . Find i=1 aiy(t . . . 1) y(N . 1) y(n + 1) y(n) . 3 7 5 Y =4 y(n) y(n . . n)]T . 1) : : : y (t . Moses. i) = y(t) + 'T (t) = a1 : : : an]T to minimize f( ) = X N This gives 2 y=6 4 y(n + 1) y(n + 2) y(N ) . n) .
Moses.. Direct Solution: For one given value of n: For k Rn+1 1 0 = 0 .Levinson–Durbin Algorithm Fast.n . {z .1 . n . . .. ... Prentice Hall. Stoica and R. . . . orderrecursive solution to YW equations 2 6 6 6 4  0 . 0 2 n 3 7 7 7 5 O(n3) ﬂops = 1 : : : n: O(n4) ﬂops Levinson–Durbin Algorithm: Exploits the Toeplitz form of Rn+1 to obtain the solutions for k : : : n in O n2 ﬂops! =1 ( ) Lecture notes to accompany Introduction to Spectral Analysis by P.0 . 1997 Slide L4–8 .. 32 7 7 7 5 } 6 6 6 4 1 n 3 7 7 7 5 2 6 6 6 4 ^ k = either r(k) or r(k).1 1 .
where x = xn : : : x1]T ~ ~ ~ Nested structure 2 Rn+2 = Thus.LevinsonDurbin Alg. Stoica and R. n 3 7 5 1 Rn+2 4 1 n = 0 3 5 2 4 Rn+1 n+1 r ~n n+1 r ~n 0 32 54 1 n n = 0 0 n 3 5 2 4 2 3 5 where r n = n+1 + ~n n Lecture notes to accompany Introduction to Spectral Analysis by P. 2 4 Rn+1 n+1 r ~n n+1 r ~n 0 3 5 2 r ~n = 6 4 . con't Relevant Properties of R: Rx = y $ Rx = y. Moses. . 1997 Slide L4–9 . Prentice Hall.
n = n ) n+1 = 0 + kn 2+1 = n + kn n = n(1 . Moses. jknj2) 2 2 n " n # " ~n 1 # Computation count: k ﬂops for the step k 2 ) n2 ﬂops !k+1 2 to determine f k k gn=1 k Slide L4–10 This is O (n2) times faster than the direct solution. 2 kn = . Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 . Prentice Hall. Stoica and R. con't 2 Rn+2 4 82 < 4 +2 : 1 n n = 0 0 n 3 5 2 4 2 3 5 2 Rn+2 4 39 = 5 2 0 ~n = 0n 2 1 n 3 5 2 4 3 5 2 3 5 Combining these gives: Rn 0 1 n + kn ~n 0 1 3 5 2 4 = 4 n + kn n 2 0 n + kn 2 n = 4 n 2 +1 3 5 0 0 Thus.LevinsonDurbin Alg.
MA Signals MA: n =0 y(t) = B (q)e(t) = e(t) + b1e(t .i!k (!) = jB(!)j k=. 1997 Slide L4–11 . Stoica and R. r(k) = 0 for jkj > m and 2 2 = m r(k)e. Prentice Hall. m) Thus. Moses.m X Lecture notes to accompany Introduction to Spectral Analysis by P. 1) + + bme(t .
Moses. with AR model order n . Stoica and R.i!k k=. ^(!) is not guaranteed to be 0 =0 Both methods are special cases of ARMA methods described below. Lecture notes to accompany Introduction to Spectral Analysis by P. Estimate fbkg and 2 and insert them in (!) = jB(!)j2 2 nonlinear estimation problem ^(!) is guaranteed to be 0 2.MA Spectrum Estimation Two main ways to Estimate (! ): 1. Prentice Hall. 1997 Slide L4–12 . Insert sample covariances (! ) = This is length X m f^(k)g in: r r(k)e.m ^BT (!) with a rectangular lag window of 2m + 1.
ARMA Signals ARMA models can represent spectra with both peaks (AR part) and valleys (MA part). k)] g = E f A(q)y(t)] A(q)y(t . A(q)y(t) = B (q)e(t) B (!) 2 = m=. 1997 Slide L4–13 .i!k k (!) = 2 A(!) jA(!)j2 P where k = E f B(q)e(t)] B(q)e(t . Stoica and R. k)] g = X n X n j =0 p=0 aj ap r(k + p .m k e. Prentice Hall. j ) Lecture notes to accompany Introduction to Spectral Analysis by P. Moses.
^(!) is not guaranteed to be 0 Lecture notes to accompany Introduction to Spectral Analysis by P. Estimate fai r(k)g in (!) = P m . Estimate fai bj A(!) nonlinear estimation problem. Prentice Hall. Moses.i!k k=.ARMA Spectrum Estimation Two Methods: 2g in (!) = 2 B(!) 2 1. Stoica and R. can use an approximate linear twostage least squares method ^(!) is guaranteed to be 0 2. 1997 Slide L4–14 .m k e jA(!)j2 linear estimation problem (the Modiﬁed YuleWalker method).
2) + = AR(1) with j kj ! 0 as k ! 1 Step 1: Approximate. 1) + 2y(t . 1) + e and its variance + ^K y(t . Prentice Hall. 1) + + K y(t . K ) ^2 = N 1 j^(t)j2 e N . Moses. K t=K +1 X Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. K ) 1a) Estimate the coefﬁcients modelling techniques. for some large K e(t) ' y(t) + 1y(t .TwoStage LeastSquares Method Assumption: The ARMA model is invertible: A e(t) = B(q) y(t) (q ) = y(t) + 1y(t . f kgK=1 by using AR k 1b) Estimate the noise sequence ^(t) = y(t) + ^1y(t . 1997 Slide L4–15 .
n) e ^(t . Prentice Hall. Note that the ai and bj coefﬁcients enter linearly in the above equation: y(t) . 1) : : : . con't fe(t)g by ^(t) in the ARMA equation. 1997 Slide L4–16 . Stoica and R.TwoStage LeastSquares Method. m)] e e = a1 : : : an b1 : : : bm]T Lecture notes to accompany Introduction to Spectral Analysis by P. ^(t) ' .y(t . 1) : : : ^(t . Moses. e A(q)y(t) ' B (q)^(t) e and obtain estimates of fai bj g by applying least squares Step 2: Replace techniques. y(t .
Prentice Hall... r f^ g . Stoica and R. # 2 = r(m + M . . . . . r(m + M ) r(m + 1) r(m + 2) 3 7 If M n. . If M > n overdetermined squares solution for ai . n + 1) r(m . least f^ g Note: For narrowband ARMA signals. Moses. i) = 0 k > m 3 7 5 " r(m) r(m + 1) . 1997 Slide L4–17 .6 4 Replace fr(k)g by f^(k)g and solve for faig. 1) : : : r(m . the accuracy of ai is often better for M > n Lecture notes to accompany Introduction to Spectral Analysis by P. n + M ) . 5 . fast Levinsontype algorithms exist for obtaining ai . . . = m + 1 ::: m + M a1 an . . : : : r(m .Modiﬁed YuleWalker Method ARMA Covariance Equation: r(k) + 2 6 4 X n In matrix form for k i=1 air(k . n + 2) . = f^ g Y W system of equations.
1997 Slide L4–18 . Stoica and R.Summary of Parametric Methods for Rational Spectra Method AR: YW or LS low lowmedium mediumhigh mediumhigh medium lowmedium No No Yes Computational Burden low Accuracy medium Guarantee ^(! ) 0 ? Yes MA: BT ARMA: MYW ARMA: 2Stage LS Use for Spectra with (narrow) peaks but no valley Broadband spectra possibly with valleys but no peaks Spectra with both peaks and (not too deep) valleys As above Lecture notes to accompany Introduction to Spectral Analysis by P. Moses. Prentice Hall.
Parametric Methods for Line Spectra — Part 1 Lecture 5 Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Prentice Hall. 1997 Slide L5–1 . Moses.
Prentice Hall. Examples: communications radar. sonar geophysical seismology ARMA model is a poor approximation Better approximation by Discrete/Line Spectrum Models 2 6(2 1) 2 6 (!) 6 (2 2) 2 6 (2 2) 3 ! Slide L5–2 . !1 !2 !3  An “Ideal” line spectrum Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses. 1997 .Line Spectra Many applications have signals with (near) sinusoidal components.
] (prevents model ambiguities) f'k g = independent rv's. power f g f g y(t) = x(t) + e(t) t = 1 2 : : : n x(t) = k ei(!k t+ k ) k=1 xk (t) X  {z } Assumptions: A1: A2: k>0 !k 2 . 1997 . superimposed in white noise of k 2. Moses. ] (realistic and mathematically convenient) A3: e(t) = circular white noise with variance 2 E fe(t)e (s)g = 2 t s E fe(t)e(s)g = 0 (can be achieved by “slow” sampling) Slide L5–3 Lecture notes to accompany Introduction to Spectral Analysis by P.Line Spectral Signal Model Signal Model: Sinusoidal components of frequencies !k and powers 2 . Stoica and R. uniformly distributed on . Prentice Hall.
k) = 2 ei!pk p j p n o Hence. e E xp(t)xj (t . for p 6= j =2 .i'j o e Z n 1 i' d' 2 = 0. !p) + 2 p Slide L5–4 Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 .i'j o = 1. for p = j e i'p e. k)g = n=1 2ei!pk + 2 k 0 p p P and (! ) = 2 X n p=1 2 (! .Covariance Function and PSD Note that: E E n i'p e.i'j o = E nei'p o E ne. r(k) = E fy(t)y (t . Moses. Prentice Hall. Stoica and R.
1997 Slide L5–5 . Moses. ^ ei!k t + residuals k Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall. Stoica and R.Parameter Estimation Estimate either: f!k k 'k gn=1 k 2 (Signal Model) f!k 2gn=1 2 k k Once (PSD Model) ^ Major Estimation Problem: f!k g f!kg are determined: ^ X f ^2g can be obtained by a least squares method from k n 2ei!pk + residuals ^ ^ r(k) = p p=1 OR: Both k and 'k can be derived by a least squares method from f^ g f^ g X y(t) = with n k= k=1 k ei'k .
Nonlinear Least Squares (NLS) Method min' g f! Let: X N k k k t=1  y(t) . . 1997 Slide L5–6 . B = eiN!1 eiN!n k 2 6 4 3 7 5 Lecture notes to accompany Introduction to Spectral Analysis by P. Moses. Prentice Hall. . X n k=1 {z F (! ') k ei(!k t+'k ) 2 } = kei'k = 1 : : : n]T Y = y(1) : : : y(N )]T ei!1 ei!n . Stoica and R. .
(B B). Prentice Hall.1B Y ] B B] . B ) = kY . Y B(B B). B k2 = .Nonlinear Least Squares (NLS) Method. Stoica and R. (B B).1B Y ^ ! Lecture notes to accompany Introduction to Spectral Analysis by P.1B Y !=^ ! and ! = arg max Y B (B B ).1B Y ] +Y Y .1B Y This gives: ^ = (B B). Moses. con't Then: F = (Y . 1997 Slide L5–7 . B ) (Y .
NLS Properties Excellent Accuracy: 6 2 (for N var (^k ) = 3 2 ! N k Example: N = 300 SNRk = 2= 2 = 30 dB k q 1) Then var (^k ) 10. ! Difﬁcult Implementation: The NLS cost function F is multimodal. Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall.5. Stoica and R. 1997 Slide L5–8 . Moses. it is difﬁcult to avoid convergence to local minima.
Unwindowed Periodogram as an Approximate NLS Method
For a single (complex) sinusoid, the maximum of the unwindowed periodogram is the NLS frequency estimate: Assume: Then:
n=1
X
B B=N
N t=1
BY=
y(t)e;i!t = Y (!)
(ﬁnite DTFT)
1 Y B (B B );1B Y = N jY (!)j2 = ^p(!) = (Unwindowed Periodogram)
So, with no approximation,
! = arg max ^p(!) ^ !
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L5–9
Unwindowed Periodogram as an Approximate NLS Method, con't
Assume: Then:
n>1
f!kgn=1 ' the ^ k
provided that
locations of the peaks of p !
^( )
n
largest
inf j!k ; !pj > 2 =N
which is the periodogram resolution limit. If better resolution desired then use a High/Super Resolution method.
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L5–10
HighOrder YuleWalker Method Recall:
y(t) = x(t) + e(t) =
k ei(!k t+'k ) +e(t)  {z } k=1 xk (t)
X
n
“Degenerate” ARMA equation for y (t):
n
(1 ; ei!k q;1)xk (t) = k ei(!kt+'k) ; ei!k ei !k(t;1)+'k] = 0
o
Let
B (q) = 1 +
4 bk q;k = A(q)A(q) k=1 A(q) = (1 ; ei!1 q;1) (1 ; ei!n q;1) A(q) = arbitrary
X
L
Then B
(q)x(t) 0 ) B (q)y(t) = B (q)e(t)
Slide L5–11
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
1997 Slide L5–12 . Stoica and R. n ^ k Lecture notes to accompany Introduction to Spectral Analysis by P.HighOrder YuleWalker Method. f!kgn=1. Moses. con't Estimation Procedure: Estimate f^igL=1 using an ARMA MYW technique b i ^( ) Roots of B q give “spurious” roots. Prentice Hall. along with L .
. Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall. 1997 Slide L5–13 . .HighOrder and Overdetermined YW Equations ARMA covariance: r (k ) + X L i=1 bir(k . r(L + 2) . r(L + M . 1) : : : r(M ) r(L + M ) 4 4 = = {z }  {z r(L) r(L + 1) This is a highorder (if L > n) and overdetermined (if M > L) system of YW equations. . . i) = 0 k > L In matrix form for k 2 6 6 6 4  = L + 1 ::: L + M 3 7 7 7 5 2 6 6 6 4 3 7 7 7 5 } : : : r(1) r(L + 1) : : : r(2) b = . . Moses. Stoica and R.
con't Fact: SVD of rank : ( )=n =U V U = (M n) with U U = In V = (n L) with V V = In = (n n). if the MinimumNorm solution b is used.1U Important property: The additional (L . diagonal and nonsingular Thus. Stoica and R. The MinimumNorm solution is b = . (U V )b = . y = . Moses.HighOrder and Overdetermined YW Equations. Lecture notes to accompany Introduction to Spectral Analysis by P.V . n) spurious zeros of B (q ) are located strictly inside the unit circle. 1997 Slide L5–14 . Prentice Hall.
Prentice Hall. . Compute ^ = . f^ g When the SNR is low. Practical Solution Let ^= ^ ^ ^ ^ but made from f^(k)g instead of fr(k)g. SVD of .HOYW Equations.1U ^ ^ ^ b ^(q) that Then !k n=1 are found from the n zeroes of B k are closest to the unit circle. this approach may give spurious frequency estimates when L > n. this is the price paid for increased accuracy when L > n. Moses.V ^ . r . V be deﬁned similarly to U . Stoica and R. Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 Slide L5–15 . V from the Let U .
Moses.Parametric Methods for Line Spectra — Part 2 Lecture 6 Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Prentice Hall. 1997 Slide L6–1 .
0 2 n 3 7 5 Lecture notes to accompany Introduction to Spectral Analysis by P. = Ax(t) + ~(t) ~ e where x(t) = x1(t) : : : xn(t)]T ~ ~(t) = e(t) : : : e(t .. m + 1)]T e 4 R = E fy(t)~ (t)g = APA + 2I ~ y Then with 2 P = E fx(t)~ (t)g = ~ x 6 4 2 1 0 . Prentice Hall. 1997 Slide L6–2 . m + 1) . . Stoica and R.i! : : : e.. Moses. 1) (for m 3 7 7 7 5 n) 2 46 y(t) = 6 ~ 6 4 y(t .The Covariance Matrix Equation Let: a(!) = 1 e.1)! ]T A = a(!1) : : : a(!n)] (m n) rank Note: Deﬁne (A) = n y(t) y(t .i(m.
1997 Slide L6–3 .Eigendecomposition of R and Its Properties R = APA + 2I (m > n) Let: 1 2 ::: m: eigenvalues of R fs1 : : : sng: orthonormal eigenvectors associated with f 1 : : : ng fg1 : : : gm.n] (m (m . Stoica and R. Moses. n)) Thus. m 7 5 S G # Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall.ng: orthonormal eigenvectors associated with f n+1 : : : m g S = s1 : : : sn] (m n) G = g1 : : : gm.. 2 R = S G] 6 4 1 3 " ..
. = = nonsingular 0 n.Eigendecomposition of R and Its Properties. con't As rank (APA ) = n: k > 2 k = 1 ::: n k = 2 k = n + 1 ::: m 0 1. Prentice Hall. since S G j j 6= 0 4 .. 0 n 3 7 5 with C (since rank S . 2 . .. 1997 . =0 A G=0 Slide L6–4 Lecture notes to accompany Introduction to Spectral Analysis by P. 2 2 6 4 3 7 5 2 6 4 Note: RS = APA S + 2S = S 1 0 . Stoica and R. Therefore.1) = AC S = A(PA S ( ) = rank(A) = n). Moses.
Moses. Prentice Hall.MUSIC Method a (!1) . f!kgn=1 are the unique solutions of k a (!)GG a(!) = 0. 1997 Slide L6–5 . Stoica and R. . A G= G=0 a (!n) ) fa(!k )gn=1 ? R(G) k 2 6 4 3 7 5 Thus. Let: 1 N y(t)~ (t) ^ ~ y R = N t=m ^^ S G = S G made from the ^ eigenvectors of R X Lecture notes to accompany Introduction to Spectral Analysis by P.
Lecture notes to accompany Introduction to Spectral Analysis by P. Here. Stoica and R.Spectral and Root MUSIC Methods Spectral MUSIC Method: f!kgn=1 = the locations of the n highest peaks of the ^ k “pseudospectrum” function: 1 ^G a(!) ! 2 . a(z) = 1 z.1)GG a(z) = 0 that are closest to the unit circle. 1997 Slide L6–6 . a (!)G ^ Root MUSIC Method: ] f!kgn=1 = the angular positions of the n roots of: ^ k ^^ aT (z.1)]T Note: Both variants of MUSIC may produce spurious frequency estimates.(m. Prentice Hall. Moses.1 : : : z.
1)^1 = 0 g no problem with spurious frequency estimates computationally simple (much) less accurate than MUSIC with m n+1 Lecture notes to accompany Introduction to Spectral Analysis by P. If: =n+1 m=n+1 Then: ^ g G = ^1. Moses.Pisarenko Method Pisarenko is a special case of MUSIC with m (the minimum possible value). ) f!kgn=1 can be found from the roots of ^ k aT (z. 1997 Slide L6–7 . Prentice Hall. Stoica and R.
MinNorm Method Goals: Reduce computational burden. Prentice Hall. R( ) Let " 1 = the vector in R(G). Uses m n (as in MUSIC). Moses. g # Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. with ﬁrst element equal ^ ^ to one. and reduce risk of false frequency estimates. 1997 Slide L6–8 . that has minimum Euclidean norm. but only one vector in G (as in Pisarenko).
1) 1 ^ g " # that are closest to the unit circle. Moses. 1997 Slide L6–9 . Stoica and R.MinNorm Method. con't Spectral MinNorm f!gn=1 = ^k the locations of the n highest peaks in the “pseudospectrum” 2 1 = a (!) 1 ^ g " # Root MinNorm f!gn=1 = ^k the angular positions of the n roots of the polynomial aT (z. Prentice Hall. Lecture notes to accompany Introduction to Spectral Analysis by P.
1 = =(1 . k k2) ) ^ = .1 g + S S .1 gm " # " # MinNorm Method: Determining g ^ Then: 1 2 R(G) ) S 1 = 0 ^ ^ ^ ^ g g ) S ^= . k k2) g Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall.S =(1 . at least.1 exists iff = k k2 6= 1 (This holds. Moses. for N 1.S(S S ).Let S ^= S g1 . g " # MinNorm solution: As: ^ = . 1997 Slide L6–10 . (S S ). Stoica and R. k k2) = (S S ) ) (S S ).) ^^ I=S S= Multiplying the above equation by gives: (1 .
i!1 0 .ESPRIT Method Let A1 = Im. determined as is uniquely = (S1S1). Prentice Hall. 1997 Slide L6–11 .1]S 2 6 4  3 7 5 Recall S = AC with jC j 6= 0. where e. let = A1D.1 0]S S2 = 0 Im.1]A Then A2 Also.i!n S1 = Im. Moses..1DC {z } So has the same eigenvalues as D .. Then S2 = A2C = A1DC = S1 C . D= 0 e.1 0]A A2 = 0 Im. Stoica and R.1S1S2 Lecture notes to accompany Introduction to Spectral Analysis by P.
1997 Slide L6–12 . Moses. ^ ^ ^ ^ The frequency estimates are found by: where f^kgn=1 are the eigenvalues of k ^ = (S1S1). arg(^k ) ^ k ESPRIT Advantages: computationally simple no extraneous frequency estimates (unlike in MUSIC or Min–Norm) accurate frequency estimates Lecture notes to accompany Introduction to Spectral Analysis by P. ﬁnd S . Stoica and R.1S1S2 ^^ ^^ f!kgn=1 = .ESPRIT Implementation From the eigendecomposition of R. Prentice Hall. then S1 and S2 .
Moses. Prentice Hall.Summary of Frequency Estimation Methods Method Periodogram Nonlinear LS YuleWalker Pisarenko MUSIC MinNorm ESPRIT Computational Accuracy / Risk for False Burden Resolution Freq Estimates small mediumhigh medium very high very high very high medium high medium small low none high high medium medium high small medium very high none Recommendation: Use Periodogram for mediumresolution applications Use ESPRIT for highresolution applications Slide L6–13 Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. 1997 .
Prentice Hall. Stoica and R. Moses. 1997 Slide L7–1 .Filter Bank Methods Lecture 7 Lecture notes to accompany Introduction to Spectral Analysis by P.
(!) by a 2. Lecture notes to accompany Introduction to Spectral Analysis by P. Parametric Approach: Parameterize ﬁnitedimensional model. Nonparametric Approach: Implicitly smooth ! !=. but 2 requires to ensure that the number of estimated values = = ) is < N . Prentice Hall.Basic Ideas Two main PSD estimation approaches: 1. ( =2 2 =1 N > 1 leads to the variability / resolution compromise associated with all nonparametric methods. Stoica and R. by assuming that ! is nearly constant over the bands f ( )g ( ) !. 1997 Slide L7–2 . Moses. !+ ] N >1 1 2 is more general than 1.
jH ( )j2 ( )d = ' 21 !+ !. Z 6 ~ (! ) @ ! (b) Z !c ^FB (!) ' 1 2 (a) . Slide L7–3 Lecture notes to accompany Introduction to Spectral Analysis by P. as well as (b) and (c). Moses. are conﬂicting.with varying ! Power Calculation Bandpass Filter c and xed bandwidth  Power in the band  Division by lter bandwidth ^(!c )  6 y y (! ) .Filter Bank Approach to Spectral Estimation H 6 (!) 1 !c ! Filtered Signal y (t) . Stoica and R. Prentice Hall. c ( )d = (') (!) (a) consistent power calculation (b) Ideal passband ﬁlter with bandwidth (c) ( ) constant on 2 ! . 2 ! + 2 ] Note that assumptions (a) and (b). !c @Q Q ! . 1997 .
Stoica and R.!) .i!k = 1 eiN (~.i!t 2 ~ ^p(~) = ! 4 N t=1 2 N 1 ~ = N y(t)ei!(N . Moses. 1 H (!) = hk e ! N ei(~. Prentice Hall.Filter Bank Interpretation of the Periodogram 1 N y(t)e.t) t=1 X X = N where ( X 1 k=0 hk y(N . 1 k=0 X 1 ei!k k = 0 : : : N . 1997 . 1 ~ hk = N otherwise 0 1 center frequency of H 3dB bandwidth of H (!) = ! ~ (!) ' 1=N Slide L7–4 Lecture notes to accompany Introduction to Spectral Analysis by P.!) . k) 2 ! .
!). and: narrow ﬁlter passband. ! 0 −5 −10 −15 dB −20 −25 −30 −35 −40 −3 −2 −1 0 1 ANGULAR FREQUENCY 2 3 Conclusion: The periodogram p ! is a ﬁlter bank PSD estimator with bandpass ﬁlter as given above. for N = 50. con't jH (!)j as a function of (~ . Stoica and R. Moses. ^( ) Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall. power calculation from only 1 sample of ﬁlter output.Filter Bank Interpretation of the Periodogram. 1997 Slide L7–5 .
Prentice Hall. more data points for the power calculation stage. Split the available sample. This approach leads to Bartlett and Welch methods. Each ﬁlter covers a small band centered on ! . Both approaches compromise bias for variance.Possible Improvements to the Filter Bank Approach 1. Stoica and R. Moses. and in fact are quite related to each other: splitting the data sample can be interpreted as a special form of windowing or ﬁltering. Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 Slide L7–6 . This “multiwindow approach” is similar to the Daniell method. Use several bandpass ﬁlters on the whole sample. and bandpass ﬁlter each subsample. 2. ~ provides several samples for power calculation.
Prentice Hall. Stoica and R.i!k = h a(!) Lecture notes to accompany Introduction to Spectral Analysis by P.i! : : : e. Moses.Capon Method Idea: Datadependent bandpass ﬁlter design. y(t . 1997 Slide L7–7 .im! ]T hk e. yF (t) = X m k=0  hk y(t . m) h y(t) ~ E jyF (t)j2 = h Rh R = E fy(t)~ (t)g ~ y {z } 7 5 {z } n o y(t) 3 H (! ) = where a X m (! ) = 1 k=0 e. = h0 h1 : : : hm] . k) 2 6 4  .
Capon Method.m t=m+1 X Lecture notes to accompany Introduction to Spectral Analysis by P. Moses. con't Capon Filter Design Problem: min(h Rh) h Solution: subject to h a(!) = 1 h0 = R. 1997 Slide L7–8 . n The Bandwidth 1 ' m+1 = (ﬁlter 1 length) Conclusion Estimate PSD as: +1 ^(!) = m^. Stoica and R.1a o The power at the ﬁlter output is: E jyF (t)j2 = h0Rh0 = 1=a (!)R.1a=a R. Prentice Hall.1a(!) which should be the power of y (t) in a passband centered on ! .1 a (!)R a(!) with N 1 ^ y(t)~ (t) ~ y R = N .
1997 Slide L7–9 . Prentice Hall. Stoica and R. but it splits the N data point sample into (N m) subsequences of length m with maximum overlap.Capon Properties m is the user parameter that controls the compromise between bias and variance: – as m increases. Moses. Capon uses one bandpass ﬁlter only. bias decreases and variance increases. Lecture notes to accompany Introduction to Spectral Analysis by P. .
jkj ^BT (!) = r(k)e.i!(t. s)e.i!k ^ k=.1a(!) Lecture notes to accompany Introduction to Spectral Analysis by P.Relation between Capon and BlackmanTukey Methods Consider ^BT (!) with Bartlett window: m m + 1 .s) = m+1 r t=0 s=0 ^ a (!)Ra(!) R = ^(i . Stoica and R.m m + 1 1 m m ^(t . 1997 Slide L7–10 . Prentice Hall. j )] ^ r = m+1 X X X Then we have ^ a (!)Ra(!) ^BT (!) = m+1 m+1 ^C (!) = ^ a (!)R. Moses.
Moses. AR Then ^C (!) = 1 1 m 1= ^AR(!) m + 1 k=0 k X Consequences: Due to the average over k. 1997 Slide L7–11 . C ! generally has worse resolution and bias properties than the AR method.Relation between Capon and AR Methods Let 2 ^k ^k (!) = ^ 2 jAk(!)j be the kth order AR PSD estimate of y (t). ^( ) ^( ) Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. C ! generally has less statistical variability than the AR PSD estimator. Due to the loworder AR terms in the average. Prentice Hall.
Stoica and R.Spatial Methods — Part 1 Lecture 8 Lecture notes to accompany Introduction to Spectral Analysis by P. Moses. 1997 Slide L8–1 . Prentice Hall.
sonar. @.@ .R . Sensor 1 @.. . Sensor 2 Sensor m Problem: Detect and locate n radiating sources by using an array of m passive sensors. @.@ . @. @. seismometers Applications: Radar. Emitted energy: Acoustic. electromagnetic. communications. Moses. mechanical Receiving sensors: Hydrophones. underwater surveillance Basic Approach: Determine energy distribution over space (thus the name “spatial spectral analysis”) Lecture notes to accompany Introduction to Spectral Analysis by P. . seismology.. @. antennas. Prentice Hall. v Source n @.. 1997 Slide L8–2 . .The Spatial Spectral Estimation Problem Source 1 vSource 2 B B B B N B cccc ccc c c c c c v @ . Stoica and R. .
) Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses. The number of sources n is known. 1997 Slide L8–3 . the array is calibrated. Prentice Hall. AOA).Simplifying Assumptions Farﬁeld sources in the same plane as the array of sensors Nondispersive wave propagation Hence: The waves are planar and the only location parameter is direction of arrival (DOA) (or angle of arrival. (We do not treat the detection problem) The sensors are linear dynamic elements with known transfer characteristics and known locations (That is.
Lecture notes to accompany Introduction to Spectral Analysis by P. Moses..g. Then the output of sensor k is yk (t) = hk (t) x(t .. background noise. etc.g. thermal noise in sensor electronics. 1997 Slide L8–4 . Prentice Hall.) hk (t) = ek (t) = Note: t 2 R (continuoustime signals). Basic Problem: Estimate the time delays known but x t unknown. Stoica and R.Array Model — Single Emitter Case x(t) = k= the signal waveform as measured at a reference point (e. k) + ek (t) ( = convolution operator). () f kg with hk(t) This is a timedelay estimation problem in the unknown input case. at the “ﬁrst” sensor) the delay between the reference point and the kth sensor the impulse response (weighting function) of sensor k “noise” at the kth sensor (e.
Narrowband Assumption Assume: The emitted signals are narrowband with known carrier frequency !c . k) ' (t) '(t . !c k ] hk (t) x(t . 1997 . k ) ' jHk (!c)j (t)cos !ct + '(t) . Then: x(t) = (t) cos !ct + '(t)] where (t) '(t) vary “slowly enough” so that (t . !c k + arg Hk (!c)] + ek(t) Slide L8–5 Lecture notes to accompany Introduction to Spectral Analysis by P. k ) ' '(t) Time delay is now ' to a phase shift !c k : x(t . Moses. k ) ' (t) cos !ct + '(t) . the kth sensor output is yk (t) = jHk (!c)j (t) cos !ct + '(t) . !c k + argfHk (!c)g] where Hk function (!) = Ffhk(t)g is the kth sensor's transfer Hence. Prentice Hall. Stoica and R.
i!c k + ek (t) where s(t) is the complex envelope of x(t).i! + ek (t)  s(t) {z }  Hk (!c) {z k c } c k yk (t) = s(t)Hk (!c) e. Stoica and R.i!ct ~ = yk(t) cos(!ct) . Moses.i!ct to obtain (t)ei (t) lowpass highpass Hence. by lowpass ﬁltering and sampling the signal yk (t)=2 = yk(t)e.!ct = (t)f ei (t) + e.Complex Signal Representation The noisefree output has the form: z(t) = (t) cos !ct + (t)] = = (t) ei !ct+ (t)] + e. Prentice Hall.i !ct+ (t)] 2 Demodulate z (t) (translate to baseband): 2z(t)e. iyk(t) sin(!ct) we get the complex representation: (for t 2 Z ) yk (t) = (t) ei'(t) jHk (!c)j ei arg H (! )] e. Lecture notes to accompany Introduction to Spectral Analysis by P.i 2!ct+ (t)] g n o  {z }  {z } Lowpass ﬁlter 2z(t)e. 1997 Slide L8–6 or .
Vector Representation for a Narrowband Source
Let
= m =
a( ) =
2 6 4
the emitter DOA the number of sensors
2 6 4
H1(!c) e;i!c
. .
1
3 7 5
Hm(!c) e;i!c m y1(t) e1(t) . . y(t) = e(t) = . . ym(t) em(t)
3 7 5 2 6 4
3 7 5
Then
y(t) = a( )s(t) + e(t)
NOTE: enters a via both For omnidirectional sensors the on .
()
f kg and fHk(!c)g. fHk(!c)g do not depend
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L8–7
Analog Processing Block Diagram
Analog processing for each receiving array element
SENSOR DEMODULATION
A/D CONVERSION


Internal Noise External Noise XXXXX ? XX ll z l  Cabling and , Filtering : ,, Incoming Transducer Signal x(t ; k )
Re yk (t)] Lowpass Re yk (t)]  !! ~ Filter A/D cos(!ct) 6 y k (t) Oscillator sin(!ct) ? ~  Im yk (t)] Lowpass Im yk (t)]  !! Filter A/D

Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997
Slide L8–8
Multiple Emitter Case Given n emitters with received signals: DOAs:
fsk(t)gn=1 k
k
Linear sensors
) y(t) = a( 1)s1(t) +
+ a( n)sn(t) + e(t)
Let
A = a( 1) : : : a( n)] (m n) s(t) = s1(t) : : : sn(t)]T (n 1)
Then, the array equation is:
y(t) = As(t) + e(t)
Use the planar wave assumption to ﬁnd the dependence of k on .
Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. Moses, Prentice Hall, 1997 Slide L8–9
1997 Slide L8–10 .Uniform Linear Arrays Source h JJ J J J J JJ J ^ J J JJ JJ+ JJ JJ 2 ] J d sin J ^ JJ J JJ 3 J J v 4 1 v ]? v d  v ppp m vSensor ULA Geometry Sensor #1 = time delay reference Time Delay for sensor k: where c = wave propagation speed d sin k = (k . Moses. 1) c Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall. Stoica and R.
Moses.i!s : : : e.i(m. Slide L8–11 d = spatial sampling period Lecture notes to accompany Introduction to Spectral Analysis by P.Spatial Frequency Let: sin 4 !s = !c d sin = 2 dc=f = 2 d sin c c = c=fc = signal wavelength a( ) = 1 e.1)!s ]T By direct analogy with the vector a(! ) made from uniform samples of a sinusoidal time series. 1997 . Stoica and R. Prentice Hall. !s = spatial frequency The function !s 7! a( ) is onetoone for j!sj $ dj sin j 1 d =2 As =2 d =2 is a spatial Shannon sampling theorem.
Moses. Stoica and R. Prentice Hall. 1997 Slide L9–1 .Spatial Methods — Part 2 Lecture 9 Lecture notes to accompany Introduction to Spectral Analysis by P.
Moses. Stoica and R. 1997 Slide L9–2 . Lecture notes to accompany Introduction to Spectral Analysis by P.Spatial Filtering Spatial ﬁltering useful for DOA discrimination (similar to frequency discrimination of timeseries ﬁltering) Nonparametric DOA estimation There is a strong analogy between temporal ﬁltering and spatial ﬁltering. Prentice Hall.
Prentice Hall.i! : : : e. Moses. 1997 Slide L9–3 . Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R.1)! ]T We can select h to enhance or attenuate signals with different frequencies ! .i(m.Analogy between Temporal and Spatial Filtering Temporal FIR Filter: yF (t) = hk u(t . m + 1)]T If u m. k) = h y(t) k=0 h = ho : : : hm.1 X (t) = ei!t then yF (t) = h a(!)] u(t)  {z } ﬁlter transfer function a(!) = 1 e.1] y(t) = u(t) : : : u(t .
Lecture notes to accompany Introduction to Spectral Analysis by P.i!c 2 : : : e. Spatial FIR Filter output: yF (t) = X m k=1 hk yk (t) = h y(t) Narrowband Wavefront: The array's (noisefree) response to a narrowband ( sinusoidal) wavefront with complex envelope s t is: () y(t) = a( )s(t) a( ) = 1 e.i!c m ]T yF (t) = h a( )] s(t)  {z } The corresponding ﬁlter output is ﬁlter transfer function We can select h to enhance or attenuate signals coming from different DOAs. Stoica and R. Moses. 1997 Slide L9–4 .Analogy between Temporal and Spatial Filtering Spatial Filter: fyk(t)gm=1 = k the “spatial samples” obtained with a sensor array. Prentice Hall.
C C B C B C B C B C B A @ .i! Be 2 B !! B B B B .i! m  e {z a( ) aa m! point ( ) qq ( ) 1 C C C C C Cs(t) C C C C C A }  ? @ ? @ @ h  1 @ R @  P h a( )]s(t)  hm. 1997 Slide L9–5 .Analogy between Temporal and Spatial Filtering u(t) = ei!t z 1 s q.B . ? qq 1 ! (Spatial sampling) (b) Spatial ﬁlter Lecture notes to accompany Introduction to Spectral Analysis by P.? B P C B C B B B . Prentice Hall. s qq h  ? 1 ?  m. Moses..i m. Stoica and R. Cu(t) C B .  e {z } ? a(!) 1 ? 0 qq ( 1) 1 (Temporal sampling) (a) Temporal ﬁlter narrowband source with DOA= Z z Z Z Z ~ aa 1reference !! h 0 0 1 B aa 2.@ 0 B 1 C h @@ C B B e. ! hm.i! C @ h a(!)]u(t) R @ C . B B @ .1 q .. B .1 ?  1 .
Moses. Here. h a 0 . where 0 j ( )j = ( ) = 25 0 30 −30 Th et a (d eg ) 60 −60 90 0 2 4 6 Magnitude 8 10 −90 Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. con't Example: The response magnitude h a of a spatial ﬁlter (or beamformer) for a 10element ULA. Prentice Hall. 1997 Slide L9–6 .Spatial Filtering.
Moses. by sweeping the ﬁlter through the DOA range of interest (“goniometer”). Prentice Hall. To locate an emitter in the ﬁeld of view. hence ﬁltering out interferences located outside the ﬁlter's beam (but possibly having the same temporal characteristics as the signal).Spatial Filtering Uses Spatial Filters can be used To pass the signal of interest only. Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. 1997 Slide L9–7 .
n o n o The (dominant) peaks of h DOAs of the sources. Moses. 1997 . and evaluate the powers of the ﬁltered signals: E jyF (t)j2 = E jh ( )y(t)j2 = h ( )Rh( ) with R = E fy (t)y (t)g.Nonparametric Spatial Methods A Filter Bank Approach to DOA estimation. Prentice Hall. ( )Rh( ) give the Slide L9–8 Lecture notes to accompany Introduction to Spectral Analysis by P. Basic Ideas Design a ﬁlter h ( ) such that for each – It passes undistorted the signal with DOA = – It attenuates all DOAs 6= Sweep the ﬁlter through the DOA range of interest. Stoica and R.
Prentice Hall. () Filter Design: min (h h) h Solution: subject to h a( ) = 1 h = a( )=a ( )a( ) = a( )=m E jyF (t)j2 = a ( )Ra( )=m2 n o Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. y t can be considered as impinging on the array from all DOAs. Moses.Beamforming Method Assume the array output is spatially white: Then: R = E fy(t)y (t)g = I E jyF (t)j2 = h h n o Hence: In direct analogy with the temporally white assumption for ﬁlter bank methods. 1997 Slide L9–9 .
1997 .Implementation of Beamforming 1 N y(t)y (t) ^ R=N t=1 X The beamforming DOA estimates are: f^kg = the locations of the n largest peaks of a Ra . pj > = wavelength array length array beamwidth Inconsistency problem: Beamforming DOA estimates are consistent if n inconsistent if n > . Stoica and R. but Slide L9–10 Lecture notes to accompany Introduction to Spectral Analysis by P. Prentice Hall. Resolution Threshold: inf j k . ( )^ ( ) This is the direct spatial analog of the BlackmanTukey periodogram. Moses. 1 = 1.
1a( ) E jyF (t)j2 = 1=a ( )R. Both Beamforming and Capon are nonparametric approaches. Stoica and R. Lecture notes to accompany Introduction to Spectral Analysis by P.1a( )=a ( )R. Prentice Hall.1a( ) n o Implementation: f^kg = the locations of the n largest peaks of ^ 1=a ( )R.Capon Method Filter design: min(h Rh) subject to h a( ) = 1 h Solution: h = R. They do not make assumptions on the covariance properties of the data (and hence do not depend on them). 1997 Slide L9–11 . Moses.1a( ): Performance: Slightly superior to Beamforming.
MINNORM and ESPRIT methods of frequency estimation can be used. Prentice Hall.Parametric Methods Assumptions: The array is described by the equation: y(t) = As(t) + e(t) The noise is spatially white and has the same power in all sensors: E fe(t)e (t)g = 2I The signal covariance matrix P = E fs(t)s (t)g is nonsingular. almost without modiﬁcation. Then: R = E fy(t)y (t)g = APA + 2I Thus: The NLS. Stoica and R. Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 Slide L9–12 . MUSIC. for DOA estimation. Moses. YW.
1A y(t) t = 1 : : : N s X Then 1 N k I . A(A A). Lecture notes to accompany Introduction to Spectral Analysis by P.1A ]Rg ^ Thus. 1997 Slide L9–13 =1 . f ^k g = arg max trf A(A A).Nonlinear Least Squares Method 1 N ky(t) . this is precisely the form of the NLS method of frequency estimation. Stoica and R. Moses. Prentice Hall.1A ]y(t) = N t=1 ^ = trf I . A(A A).1A ]Rg X f kg For N . As(t)k2 min(t)g N f k g fs t=1 X  Minimizing f over s gives f ( s) {z } ^(t) = (A A).1A ]y(t)k2 f ( ^) = N s t=1 1 N y (t) I . A(A A).
Moses.Nonlinear Least Squares Method Properties of NLS: Performance: high Computational complexity: high Main drawback: need for multidimensional search. Prentice Hall. Lecture notes to accompany Introduction to Spectral Analysis by P. Stoica and R. 1997 Slide L9–14 .
Prentice Hall.n 0 V2 g L n M . Stoica and R.= y (M L) Also assume: M > n L > n ( ) m = M + L > 2n) rank ~ (A) = rank(A) = n " # " # (. = U1 U2 ] 0 n n 0 V1 g n . Moses.n ~ ~ Properties: A V2 = 0 V1 2 R(A) Then: rank {z} {z} Lecture notes to accompany Introduction to Spectral Analysis by P.YuleWalker Method e y(t) = y(t) = A s(t) + ~(t) ~ y(t) ~ e(t) A " # " # " # Assume: Then: E fe(t)~ (t)g = 0 e 4 E fy(t)~ (t)g = AP A ~ . 1997 Slide L9–15 .) = n. and the SVD of . is .
is the “array transfer vector” for y(t) a ~ at DOA ^ V2 is deﬁned similarly to V2. Stoica and R.YWMUSIC DOA Estimator f^kg = where the n largest peaks of 1=~ ( )V2V2 ~( ) a ^^a ~( ). using 1 N y(t)~ (t) ^ . Moses.=N y t=1 X Properties: Computational complexity: medium Performance: satisfactory if m 2n Main advantages: – weak assumption on e t – the subarray A need not be calibrated f ( )g Lecture notes to accompany Introduction to Spectral Analysis by P. (L 1). Prentice Hall. 1997 Slide L9–16 .
if rank (P ) = rank(E fs(t)s (t)g) < n Modiﬁcations that apply in the coherent case exist. Stoica and R. Moses. Prentice Hall.MUSIC and MinNorm Methods Both MUSIC and MinNorm methods for frequency estimation apply with only minor modiﬁcations to the DOA estimation problem. Spectral forms of MUSIC and MinNorm can be used for arbitrary arrays Root forms can be used only with ULAs MUSIC and MinNorm break down if the source signals are coherent. Lecture notes to accompany Introduction to Spectral Analysis by P. 1997 Slide L9–17 . that is.
i!c ( n) ( ) = the time delay from subarray 1 to 2 6 4 3 7 5 subarray 2 for a signal with DOA = : where d is the subarray separation and is measured from the perpendicular to the subarray displacement vector. Stoica and R.ESPRIT Method Assumption: The array is made from two identical subarrays separated by a known displacement vector. Let m = # sensors in each subarray A1 = Im 0]A (transfer matrix of subarray 1) A2 = 0 Im]A (transfer matrix of subarray 2) Then A2 = A1D. D= 0 e.i!c ( 1) 0 .. 1997 Slide L9–18 ( ) = d sin( )=c . Lecture notes to accompany Introduction to Spectral Analysis by P. where e. Prentice Hall. Moses..
Prentice Hall.1 A2 = 0 Im.1 0]A = .1]A Lecture notes to accompany Introduction to Spectral Analysis by P.ESPRIT Method. Moses. con't ESPRIT Scenario source θ known subarray 1 displac ement v ector subarray 2 Properties: Requires special array geometry Computationally efﬁcient No risk of spurious DOA estimates Does not require array calibration Note: For a ULA.1 A1 = Im.1 . 1997 Slide L9–19 . so m m and . the two subarrays are often the ﬁrst m and last m array elements. Stoica and R.