Scribd
Upload
27 views

% Ex. 1 % % QPSK % Uncorrelated Rayleigh % ML Detection % Spatial Multiplexing % % Comparing The Optimal ML Decoder With The Simple ZF Decoder

This document simulates and compares the performance of maximum likelihood (ML) detection and zero-forcing (ZF) detection for a spatial multiplexing system with 2 transmit antennas and 2 receive antennas over uncorrelated Rayleigh fading channels. It generates QPSK symbols, transmits them over simulated fading channels with additive white Gaussian noise, and detects the symbols using both ML detection (which searches all possible symbol combinations) and ZF detection (which simply inverts the channel). It calculates the symbol error rates for each detector across a range of SNR values and plots the results.

Uploaded by

Charu Shree
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
27 views

% Ex. 1 % % QPSK % Uncorrelated Rayleigh % ML Detection % Spatial Multiplexing % % Comparing The Optimal ML Decoder With The Simple ZF Decoder

This document simulates and compares the performance of maximum likelihood (ML) detection and zero-forcing (ZF) detection for a spatial multiplexing system with 2 transmit antennas and 2 receive antennas over uncorrelated Rayleigh fading channels. It generates QPSK symbols, transmits them over simulated fading channels with additive white Gaussian noise, and detects the symbols using both ML detection (which searches all possible symbol combinations) and ZF detection (which simply inverts the channel). It calculates the symbol error rates for each detector across a range of SNR values and plots the results.

Uploaded by

Charu Shree
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

%

Ex.
1
%
% QPSK
% Uncorrelated Rayleigh
% ML detection
% Spatial Multiplexing
%
% Comparing the optimal ML decoder with the simple ZF decoder
clear;
close all;
Nsym = 1e6; % Symbols
EsNo_db = 0:45;
snr = 10.^(EsNo_db/10);
Nst = 2; % number of streams
s = signc(randcn(Nst, Nsym));
Es = 2;
lstr = {};
Nrx = 2;
Ntx = 2;
n = randcn(Nrx, Nsym);
H = randcn(Nrx, Ntx, Nsym);
% generate all possible symbol combinations (4x4)
qpsk = [ 1+1j 1-1j -1-1j -1+1j ]; % all 4 QPSK options
s_test = zeros(2, 0);
for i=1:length(qpsk)
for j=1:length(qpsk)
s_test(:,end+1) = [qpsk(i) qpsk(j)].';
end
end
for i=1:length(snr)
fprintf('.');

s_hat_ml = zeros(Nst, Nsym);


s_hat_zf = zeros(Nst, Nsym);
d = zeros(1, 16);

for k=1:Nsym
Hk = H(:,:,k);
r = Hk*s(:,k)/sqrt(Ntx) + n(:, k)/sqrt(snr(i)/Es);
% Optimal ML receiver tests all posible s1, s2
% (compute the distance d -- then minimize it)
n_hat = Hk*s_test/sqrt(Ntx) - repmat(r, 1, 16);
d = sum(abs(n_hat).^2, 1);

[dmin, sym_index] = min(d);


s_hat_ml(:,k) = s_test(:,sym_index);

% Suboptimal zero-forcing equalizer just inverts the channel


s_hat_zf(:,k) = signc( pinv(Hk)*r );

end
P_err_zf(i) = mean(s_hat_zf(:) ~= s(:));
P_err_ml(i) = mean(s_hat_ml(:) ~= s(:));
end
EsNo_db_sm = EsNo_db;
save sm P_err_ml P_err_zf EsNo_db_sm
semilogy(EsNo_db_sm, P_err_zf, 'r--', ...
EsNo_db_sm, P_err_ml, 'b-.');
title('Spatial Multiplexing');
ylabel('Pr(Symbol Error)')
xlabel('E_s/N_0 [dB]');
legend('SM 2x2 (ML)', 'SM 2x2 (ZF)');
grid on;

You might also like