CHAPTER 7 |-spectrum, Problem 7.1 (a) The PN sequence length is (b) The chip duration is tT, = 1s -01ns 107 (c) The period of the PN sequence is T=NT, 15x 0.1 = 15 ns Problem 7.2 Shift Shift-register Modulo-2 Shift-register number contents adder output output 0 1000 1 0100 0+0=0 0 2 0010 0+0=0 0 3 1001 0 4 1100 O+l=1 1 5 0110 0+0=0 0 6 1011 1+0 0 7 0101 1+1=0 1 8 1010 O+1=1 1 4019 1101 14021 0 10 1110 O+1=1 1 ll 1111 1+0=1 0 12 0111 14+1=0 1 13 0011 1+1=0 1 14 0001 14+1=0 1 15 1000 O+1=1 a The output sequence is therefore 11 000100110101111 0001 one period Problem 7.3 (a) From both Table 7.2a and Table 7.2b we note the following: Balance property: Number of 1s in one period = 16 Number of 0s in one period = 15 Hence, the number of 1s exceeds the number of 0s by one. (b) Run property: In both Tables 7.2a and 7 2b, we count a total of 8 runs of 1s and a total of 8 runs of 0s, Moreover, we note the following: Runs of length 1: 4 Runs of length 2: 2 Runs of length 3: 1 (©) Autocorrelation function: olHence, we have (not to seale) 6; 403Problem 7.4 Se L—{_s |+4 Ls } +4 Table 1 Feedback State of Feedback- Output symbol Bor eEHoOnHooHnHeOHOoOHOHH HHH shift register 100000 110000 111000 111101 ee ee 111111 011111 101111 010111 101011 010101 101010 110101 011010 001101 100110 110011 011001 101100 110110 111011 011101 101110 symbol BE OoOH HOH OHOHH HHH HOO OOO 404Table 1 continued Feedback symbol . COP OH C OPH HH OH CCOHHH OOH OOH ON State of feedback- shift register 110111 011011 101101 010110 001011 100101 010010 001001 100100 110010 111001 011100 001110 000111 100011 010001 101000 110100 111010 111101 011110 001111 100111 010011 101001 010100 001010 Output symbol Bee enon CoCOH HH OOH OC OH OHH OHHH O 405Table 1 continued Feedback State of feedback- Output symbol _shift register symbol 000101 100010 110001 011000 001100 000110 000011 100001 010000 001000 000100 000010 000001 100000 HeccooHocooHHS HeoocoHooecoHoHeE 406Table 2 6 [—¥ aR Feedback State of feedback- Output symbol CH HH HHH OH OHHH OCCOHRHOCOHHH ON shift register 100000 110000 011000 101100 110110 111011 011101 001110 100111 110011 011001 001100 000110 100011 110001 111000 011100 101110 010111 101011 110101 111010 111101 111110 111111 [Ue Ue Ge ia es symbol HOH OHHH OSC OHH OCH HH OHH COO O 407Table 2 continued Feedback State of feedback- Output symbol __ shift register symbol 101111 110111 011011 101101 010110 001011 000101 100010 010001 001000 000100 000010 100001 010000 101000 110100 011010 001101 100110 010011 101001 010100 101010 010101 001010 COP SCH OH CORN SH CC OCCH OCC OH OHH HROoOHH OH OSCO OH SCC OH OHH OHH HETable 3 Feedback symbol COoOPHoCeC OH OH OH OOH H HH HHO ¥ State of feed-back shift register 100000 010000 101000 110100 111010 111101 111110 111111 011111 001111 100111 010011 101001 010100 101010 010101 001010 000101 100010 110001 011000 001100 aa Output symbol CHOH CHOC OHHH HHH OH OSC OOO 409 {os oHTable 3 continued Feedback State of feedback- Output symbol __ shift register symbol 100110 110011 111001 111100 011110 101111 110111 111011 011101 101110 010111 101011 110101 011010 101101 010110 001011 100101 110010 011001 101100 110110 011011 001101 000110 SSCHH OHH OOH OHH OH OHHH OHM HE BPR OCH OMH OH OHHH OM M HHO OHHOSTable 3 continued Feedback State of feedback- Output symbol _ shift register symbol 1 100011 0 010001 0 001000 1 100100 0 010010 0 001001 0 000100 0 000010 1 100001 1 110000 1 111000 0 011100 0 001110 0 000111 0 000011 0 000001 a 100000 Hen ScocoecHoOOoHOoOoHHOTable 3 continued Feedback State of feedback Output symbol _shift register symbol 100101 010010 001001 100100 110010 111001 111100 011110 oo1111 000111 000011 000001 100000 HecoeooHH HH OOH HH HH oOoOH OC OHOHO 412State of feedback- Output shift register symbol Initial stat 10000 01000 00100 00010 00001 11101 10011 10100 01010 00101 11111 10010 01001 11001 10001 10101 10111 10110 01011 11000 01100 00110 00011 11100 01110 00111 11110 01111 11010 01101 11011 10000 BH oOHOHCOH OCC OH OHH HH HOH HOC OH HH OCC O 413The 31-element code generated by the scheme shown in Fig. P9.2 is exactly the same as that described in Table 9.2b. Note, however, the code described in Tab.e 9.2b appears in reversed order to that described in the above table; this reversal is clearly of a trivial nature. 414Problem 7.6 (a) The modulo-2 sum of b(t) and c(t), on a pulse-by-pulse basis, is as follows bw 1 e(t) 0 0 1 1f1 0 (b) If symbol 0 is represented by a sinusoid of zero phase shift, and symbol 1 is represented by a sinusoid of 180° phase shift, the output of the modulo-2 adder takes on the same form as that described in Table 7.3 of the text. Problem 7.7 itt) = V2T cos(2nfyt + 8) The basis functions are ato = cos(2nf,t), kT, St < (k+DT, c 0, otherwise Ac = sin(2nfct), kT, oO Ld sw (Ginary dats) Baseband ‘ler co) ‘The DS/QPSK modulated signal is |E |E x(t) (1)cos(2n f.1) + Feots)sin(2nf.t) where eft) = {¢9, (0, €4, (ss Cyn, CE} and Colt) = {ep, (4.61, ols ey-4, of} denote the spreading sequences for 0 <1 < T,, which are applied to the in-phase and quadrature channels of the modulator. Consider the following set of orthonormal basis functions: os(2nft), kT S1S(k+1)T, otherwise 418i [Fsmcns.n, ) AT ,StS (k+1)T, 0, otherwise Go, = where T, is the chip duration; & = 0, 1, 2, «++, N-1,and N= TIT,, that is, N is the number of chips per bit ‘The DS/QPSK modulated signal can be written as follows (using the set of basis functions): T 2 TE [2 a [Pooenranetns oF Fonens.ncac [25 ze. 190, 4, {2 Xo, 1% eg ‘The channel output at the receiving end of the system has the following form A(t) = s(t) + f(D) where j(t) denotes the interference signal. We may express the interference signal using the 2N-dimensional basis functions as follows: wa xa FO = Yer ie, + Dieg. (eg (0 k= k=O " where (EN, Jin, "HO%, (D4 Jeng (kstyr jean = Sur, HOMeg Oat k= OLN The average power of the interferer is given by 419Assuming that the power is equally distributed between the in-phase and quadrature components: N-1) Sees ee L Matis sie ef) There are two stages of demodulation. First, the received signal x(7) is multiplied by the despreading sequences et) and co(0), yielding u(t) uo(t) Feos(2m ft) £ exneo(Esint2n7.0 +e(O() Esnensn t+ eqhne so) Boosters. + eo(t)i(t) 420The second terms in the right-hand side of uj(t) and ug(t) are filtered by the bandpass filters, and the BPSK demodulators recover estimates of their respective binary sequences. Finally, the multiplexer reconstructs the original binary data stream, Proc: it The signal-to-noise ratio at the output of the receiver is (SNR), = Mstantaneous peak signal _power ‘The signal-to-noise ratio at the input of the coherent receiver is average input-signal power (SNR), average interferer power We may therefore write (SNR) oi 2: (r oe, GAR, = totes, (7) = 3+ 10log, | 7) The QPSK processing gain= T/T, That is, PGaesx = 21PGypsx] Solving for the antenna aperture: ziProblem 7.12 ‘The processing gain (PG) is PG Hence, expressed in decibels, PG = 101og1920 = 26 db Problem 7.13 The processing gain is PG =4x4 = 16 Hence, in decibels, PG = 101ogi916 = 124B 422Problem 7.13 Matlab codes % Problem 7.13(a), CS: Haykin 4% Generating 63-chip PN sequences % polynomial (x) = x°6+x+1 4% polynomial2(x) = x°6 + x°6 + x24 x44 4% Mathini Sellathurai, 10.08.1999 % polynomials polt={t 0000143; polz[1 100414]; % chip size 35 % generating the PN sequence pnseqi = PNseq(pol1); pnseq? = Pliseq(pol2); % mapping antipodal signals (0-->-1, 1-->4) u=2epnseqi~1; spnseq?-14% autocorrelation of pnseqi [eorrf]=pn_corr(u, u, M) % prints plot(~61:62,corrt(2:126)); axis([-62, 62,-10, 80]) xlabel(’ Delay \tau’) ylabel(’ Autocorrelation function R_{c}(\tau)’) pause Yautocorrelation of pnseq? [eorr#]=pn_eorr(v, v, N) % prints plot (~61:62, corrf(2:125)); axis([-62, 62,-10, 80]) xlabel(? Delay \tau’) ylabel(? Autocorrelation function R_{ch(\tau)’) pause 4% cross correlation of pnseqi, pnseq? [eucorr]=pn_corr(u, v, W) % prints plot (-61:62,c_corr(2:426)); axis([-62, 62,~20, 20]) xlabel(’ Delay \tan’) ylabel(? Cross-correlation function R{Ji}(\tau)’) 4244% Problem 7.13 (b), CS: Haykin 4% Generating 63-chip PN sequences % polynomiali(x) =x6+x+4 % polynomial2(x) = x6 + x6 + x24 x41 % Mathini Sellathurai, 10.05.1999 % polynomials polis{i 1100141; pola[t 1001147; % chip size N63; % generating the PN sequence pnseqi = PNseq(pol1); Pnseq? = PNseq(pol2); % mapping antipodal signals (0-->-1, u=2+pnseqi-1; v=2epnseq2-1} 4% autocorrelation of pnseqi (corr#]=pn_corr(u, u, MN) % prints plot(-61:62,corrt(2:126)); axis([-62, 62,-10, 80]) xlabel(’ Delay \tau’) ylabel(’ Autocorrelation function R{c}(\tau)’) pause autocorrelation of pnseq2 Ceorrf}=pncorr(v, v, N) 4 prints plot(-61:62,corré(2:125)); axis([-62, 62,-10, 80]) xlabel(? Delay \tau’) ylabel(? Autocorrelation function R_{c}(\tau)") pause % cross correlation of pneegi, pnseq? (e_corr]=pn_corr(u, v, ND % prints 425plot(~61:62,c_corr(2:426)); axis([-62, 62,-20, 20)) Habel(? Delay \tau’) ylabel(’ Cross-correlation function R{ji}(\tau)’) 426function x = Plseq(p) 4% Linear shift register for generating PN sequence of polynomial p % used for problens 7.13, 7.14 of CS: Haykin % Mathini Sellathurai, 10.08.1999 Length(p) - 1; % order of the polynomial fLipir(p); (1 zeros(1, N-1)]; 45 " P x for i= 1: ne(2-N- 1) x(4) = XC); X= [X(2:m) p(t) * rem(sum(p(1:W) .* x(4:N)), 20]; ond 427o function [corrf]=pn_corr(u, v, N) % funtion to compute the autocorreation/ cross-correlation % function of two PN sequences % used in problem 7.13, 7.14, CS: Haykin 4 Mathini Sellathurai, 10 june 1999. for m=0:11 shitted_u=[u(m+1:N) u(1:m)]; corr(m+1)=(sum(v.#shifted_u)); if (abs(corr)>max_cross_corr) max_cross_corr=abs(corr) end end 428Answer to Problem 7.13 180 pe Figure: Autocorrelation function of [6,5,2,1],6,1) 429eo = = obs 3 we 70 Figure 2: Cross-correlation function of [6,5,2,1] 6,1] beiays ire J: Cross-correlation function of (6,5,2,1]{8,5,4,1] 430Problem 7.14 Matlab codes 4% Problem 7.14 (a), CS: Haykin 1 Generating 31-chip PN sequences % polynomiali(z) = x76 + x°2 +1 % polynomial2(x) = x75 + x73 +4 % Mathini Sellathurai, 10.05.1999 % polynomials poli=[10010 17; 101004); % chip size N31; 4% generating the PN sequence pnseqi = PNseq(polt); pnseq? = PNseq(pol2): 4% mapping antipodal signals (0-->-1, ue2epnsegi-1 v-2epnseq2-1 % cross correlation of pnseqi, pnseq2 [e_corr]=pn_corr(u, v, N) % prints plot(-30:31,c_corr); axis([-30, 31-18, 18]) xlabel(? Delay \tau’) ylabel(? Cross~correlation function R_{ji}(\tau)*) 4314% Problem 7.14 (b), CS: Haykin 4% Generating 63-chip PN sequences 4% polynomiali(x) = x75 + x73 +4 4% polynomial2(x) = x°6 + x74 + De KHL 4% Mathini Sellathurai, 10.05.1999 4% polynonials % chip size N31; 4 generating the PN sequence pnseqi = Piseq(polt. phseq? = Pliseq(pol2); % mapping antipodal signals (0-->-1, epnseqi-1; epneeq2-1; 4% cross correlation of pnseqi, pneeq? (c_corrl=pn_corr(u, v, ¥) % prints plot(-30:31,c_corr); axis([-30, 31,-10, 10]) xlabel(? Delay \tau’) ylabel(? Cross-correlation function R{ji}(\tau)’)% Problem 7.14 (c), CS: Haykin % Generating 63-chip PN sequences % polynomiali(x) = x°6 + x-4 + x-3tt % polynomial2(x) = x6 tx4+x2+x+1 % Mathini Sellathurai, 10.05.1999 % polynomials 1111042; 104110; % chip size N34; 4% generating the PN sequence pnseqi = PNseq(polt); phseq? = PNseq(pol2); % mapping antipodal signals (0- ue2¢pnseqi-1; v=24pnseq?-1; % cross correlation of pnseqi, pnseq2 {e_corr]=pn_corr(u, v, N) % prints plot (-30:31,c_corr); axis([-30, 30,-10, 10]) xlabel(’ Delay \tau’) ylabel(? Cross-correlation function R_{ji}(\tau)’) 433Answer to Problem 7.14 Casco Figure {: Cross-correlation function of [5,3}[5.2] 434lint . Delay Figure 3: Cross-correlation function of {6,5.2,1],[6,5,4.1] 435