In Matlab Digital Comm

The following example codes are provided for BPSK, need to modify it to work with QAM

%bpSK_step.m Eb=1; EbN0dB=2 pause

EbN0=10.^(EbN0dB/10); N0=Eb/EbN0 pause

% Monte-Carlo verification Num_of_bits=20; bits=myrandint(1,Num_of_bits,2) grid on stem(bits) axis([1 20 0 1.5]); pause

S=(2*bits-1) stem(S) axis([1 20 -1.5 1.5]); pause

N=sqrt(N0/2)*randn(1,Num_of_bits) stem(N) axis([1 20 -2 2]); pause

X=S+N stem(X) axis([1 20 -2 2]); pause

bits_decoded= (sign(X)+1)/2 stem(bits_decoded) axis([1 20 0 1.5]); pause

fprintf('compare bits with bits_decoded'); Errors_count=sum(bits-bits_decoded~=0) bits bits_decoded pause

Pe_simu=Errors_count/Num_of_bits

if 1 ==0 figure(1) stem(S) axis([1 20 -2 2]); grid on

figure(2) stem(N) axis([1 20 -2 2]); grid on

figure(3) stem(X) axis([1 20 -2 2]); grid on end

Eb=1; EbN0dB=[1:11];

% Theoretical computation for i=1:length(EbN0dB) EbN0=10.^(EbN0dB(i)/10); N0=Eb/EbN0; Pe_th(i)=Qfunc(sqrt(2*EbN0)); end

% Monte-Carlo verification Num_of_bits=10^6; bits=myrandint(1,Num_of_bits,2); % constellation points at +1, -1 S=2*bits-1;

for i=1:length(EbN0dB) EbN0=10.^(EbN0dB(i)/10); N0=Eb/EbN0;

N=sqrt(N0/2)*randn(1,Num_of_bits); X=S+N; bits_decoded= (sign(X)+1)/2; Errors_count=sum(bits-bits_decoded~=0); Pe_simu(i)=Errors_count/Num_of_bits; end

figure(1) semilogy(EbN0dB,Pe_th,'-rs'); hold on semilogy(EbN0dB,Pe_simu,'-b*'); hold on hold off set(gca,'ylim',[1e-6,1e-1]) title('Bit Error Rate for BPSK modulation'); xlabel('Eb/N0 in dB'); ylabel('Bit Error Rate'); legend('Theory','Simulation'); grid

%function y=Qfunc(x) %y=.5*erfc(x/sqrt(2));

This project is to compare the theoretical symbol error rate (SER) expression with simulated results, the union bound, and the approximate SER expression based on nearest neighbor rules. In the first figure, plot the following four curves: (a) the theoretical SER versus E_s/N_0 (in decibels) for 16-QAM. (b) the simulated SER versus E_s/N_0 (in decibels) for16-QAM. (c) the union bound on SER versus E_s/N_0 (in decibels) for 16-QAM. (d) the approximate SER versus E_s/N_0 (in decibels) for 16-QAM. Assume that Gray mapping is used to map bits to symbols. In second figure, plot the following four curves: (a) the simulated bit error rate (BER) versus E_b/N_0 (in decibels) for 16-QAM. (b) the approximate BER (based on nearest neighbor rule) versus E_b/N_0 (in decibels) for 16-QAM. This project is to compare the theoretical symbol error rate (SER) expression with simulated results, the union bound, and the approximate SER expression based on nearest neighbor rules. In the first figure, plot the following four curves: (a) the theoretical SER versus E_s/N_0 (in decibels) for 16-QAM. (b) the simulated SER versus E_s/N_0 (in decibels) for16-QAM. (c) the union bound on SER versus E_s/N_0 (in decibels) for 16-QAM. (d) the approximate SER versus E_s/N_0 (in decibels) for 16-QAM. Assume that Gray mapping is used to map bits to symbols. In second figure, plot the following four curves: (a) the simulated bit error rate (BER) versus E_b/N_0 (in decibels) for 16-QAM. (b) the approximate BER (based on nearest neighbor rule) versus E_b/N_0 (in decibels) for 16-QAM