Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

[SOLVED] LS and MMSE channel estimation for ofdm

Status
Not open for further replies.

Romance42932

Newbie level 4
Newbie level 4
Joined
Feb 5, 2013
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Visit site
Activity points
1,353
hi...
i want to pass tx (QPSK) signal thru a awgn channel.... let say i have provided snr of 30dB and channel pdp with L=4(tap length) is exp(-n/5) where n is from(0 to 4)...
so i want to create channel with say length 5 with noise which satisfies given snr in time domain.... so that y = hx + n then i wanna do channel estimation in freq domain Y = HX + N.....
tell me also what is the equivalent variance of that noise in freq domain? (for N).... i am doing this code.... i dont know what is wrong in it...

i am using pilots as +1 and -1 ......i am trying find H values on all subcarriers .....not only on pilot....is that make any difference


%%%%%%%%%%CODE%%%%%%%%%%%%%%%
clear all
clc
sub_car = 128; % number of data subcarriers
nbitpersym = 128;
len_fft = 128;
t_pilot = 128/8;
pilotinterval = 8;
nsym = 10^4; % number of symbols
%Create a QPSK modulator System object with bits as inputs and Gray-coded signal constellation
h = comm.QPSKModulator('PhaseOffset',pi/16);
% change the phase offset to avoid same types of values
% Generating data
t_data=randint(nbitpersym,1);
% modulating data
mod_data = step(h,t_data);
% pilot insertion
mod_data(1:2*pilotinterval:end)= 1; % pilot is taking as +1 or -1
mod_data(9:2*pilotinterval:end)= -1;
IFFT_data = ifft(mod_data.',len_fft);
a = IFFT_data([1:8:end]);
% addition cyclic prefix
cyclic_add_data = [];
cyclic_add_data = [IFFT_data:),[113:128]) IFFT_data];
noise_power = [];
noise_var = [];
pdp = exp(-(0:5)/5);
snr = 30; % snr in dB
% for i= 1:length(snr)
noise_power = 2./10.^(snr./10);
noise_var = sqrt(noise_power);
% end
pdp = exp(-(0:4)/5);
hx = pdp*randn(1,5)';
channel = hx + sqrt(noise_var)*randn(1);
%passing thru the channel
rec_sig = conv(cyclic_add_data,channel);
%discarding 149-144 = 5 to get actual signal
rec_sig = [rec_sig:),[6:149])];
%cyclic prefix removal
Y = [rec_sig:),[17:144])];
% freq domain transform
FFT_data =fft(Y.').';
% demodulation
z= comm.QPSKDemodulator('PhaseOffset',pi/16);
demod_data = step(z,FFT_data');
% [no_of_error,ratio] = biterr(t_data,demod_data);
X=zeros(128,128);
for l = 1:1:128
X(l,l)=IFFT_data(l);
end
H = Y*inv(X);
G = H';
H = fft(G)
u=rand(128,128);
F=fft(u)*inv(u);% 'F' is the twiddle factor matrix..
gg=zeros(128,128);
for i=1:128
gg(i,i)=G(i);
end
gg_myu = sum(gg, 1)/128;
gg_mid = gg - gg_myu(ones(128,1),:);
sum_gg_mid= sum(gg_mid, 1);
Rgg = (gg_mid' * gg_mid- (sum_gg_mid' * sum_gg_mid) / 128) / (128 - 1);
for m=1:200

for n=1:6.0

SNR_send=5*n;
XFG=X*H;
n1=ones(128,1);
n1=n1*0.000000000000000001i;
noise=awgn(n1,SNR_send);
variance=var(noise);
N=fft(noise);
Y=XFG+N;
%the mean squared error for the LS estimator..
mean_squared_error_ls=LS_MSE_calc(X,H,Y);
%the mean squared error for the MMSE estimator..
mean_squared_error_mmse=MMSE_MSE_calc(X,H,Y,Rgg,variance);
SNR(n)=SNR_send;
mmse_mse(m,n)=mean_squared_error_mmse;
ls_mse(m,n)=mean_squared_error_ls;
end;

end;
ls_mse
mmse_mse
mmse_mse_ave=mean(mmse_mse)
ls_mse_ave=mean(ls_mse)
%plotting
semilogy(SNR,mmse_mse_ave,'k-');
grid on;
xlabel('SNR in DB');
ylabel('mean squared error');
title('PLOT OF SNR V/S MSE FOR AN OFDM SYSTEM WITH MMSE/LS ESTIMATOR BASED RECEIVERS');

hold on;
semilogy(SNR,ls_mse_ave,'b*');
semilogy(SNR,ls_mse_ave,'b-');
semilogy(SNR,mmse_mse_ave,'kv');
grid on;
xlabel('SNR in DB');
ylabel('mean squared error');
title('PLOT OF SNR V/S MSE FOR AN OFDM SYSTEM WITH MMSE/LS ESTIMATOR BASED RECEIVERS');

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%Function for LS estimation code%%%%%%5

function ms_error=LS_MSE_calc(X,H,Y);
Hls =(inv(X)) * Y;

ms_error_mat=mean(((abs(H-Hls))/abs(H)).^2);
for i=1:128
if(ms_error_mat(i)~=0)
ms_error=ms_error_mat(i);
end
end


%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%function for mmse code%%%%%%%%%



function ms_error=MMSE_MSE_calc(X,H,Y,Rgg,variance);
%Hmmse=F*Rgg*inv(Rgy)*Y;
u=rand(128,128);
F=fft(u)*inv(u);%The 64 X 64 twiddle factor matrix..
I=eye(128,128);
Rgy=Rgg * F'* X';
Ryy=X * F * Rgg * F' *X' + variance * I;
for i=1:64
yy(i,i)=Y(i);
end
Gmmse=Rgy * inv(Ryy)* Y;
Hmmse=fft(Gmmse);

ms_error_mat=mean(((abs(H)-abs(Hmmse))/abs(H)).^2);
for i=1:128
if(ms_error_mat(i)~=0)
ms_error=ms_error_mat(i);
end
end
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top