?? qam256.m
字號:
%*************************************************
% 256 QAM EQUALIZER TESTING FILE *
%*************************************************
% *
% DATE: 5/28/98; *
% AUTHOR: QIN ZHANG *
% *
% FILE NAME: EQUA_256.M (12-BIT 256 QAM) *
% C:QAM\SIMULATION *
% transmitter alpha=.12 receiver alpha=.12 *
%*************************************************
%
% Two Channel Third Order Simulation;
%path(path,'c:\matlab\mat_apl');
%path(path,'c:\matlab\mg');
%path(path,'c:\matlab\data');
clear;
fs =5e6*4; %5.304e6*4; %42.884296e6/2; % 20.227640MHz sampling frequency;
f_symbol=fs/4; % 10.113820MSPS
taps_t=5*2*2*2*2*2; % 160-tap filter;
alpha_t=0.1865; %alpha_t=0.1865 % Transmiter SRRC filter alpha;
taps_r=taps_t; % receiver filter taps;
alpha_r=alpha_t; % ideal: 0.1865 % receiver fitler alpha;
n_filter=taps_t*2-1; % raised cosine filter taps;
n_bit_rc =16; % raised cosine filter coefficient bit;
n=1024*4*2; % random data size;
n_demo_filter=6; % demodulation low pass filter taps;
f_cutoff_demo=5e6; % demodulation low pass cutoff frequency;
n_bit_da=16; % D/A converter resolution;
n_bit_sin=16; % Sine and Cosine Coefficient;
ph_error=0; % carrier phase error;
ny_full=1; % full nyquist filter; 0 partial
ny_n=10; % partial nyquist filter, 8 taps on left and right from center tap;
%*****************************************************************
% 1) Generate Square Root Raised Cosine Filter Coefficients *
%*****************************************************************
t=1:taps_t;
t=(t-floor(taps_t/2)-1+1e-10)/fs;
hrc_t=(sin(pi*t*f_symbol*(1-alpha_t)) + 4*alpha_t*t*f_symbol.*...
cos(pi*t*f_symbol*(1+alpha_t)) )./...
( pi*t*f_symbol.*(1-(4*alpha_t*t*f_symbol).^2));
%hrc_t=hrc_t.*blackman(taps_t)';
hrc_t=hrc_t.*kaiser(taps_t,6)';
plot(t, hrc_t);
grid;
title('hrc_t impulse response');
pause;
hrc_r=hrc_t;
hrc_t=hrc_t/sum(hrc_t);
hrc_r=hrc_r/sum(hrc_r);
maxmin=max(abs(hrc_t));
module=1/(2^(n_bit_rc-1)-1)*maxmin; % define LSB;
hrc_t_i=round(hrc_t/module);
% transform the fitler to polyphase;
hrc_t_4=reshape(hrc_t,4,taps_t/4);
hrc_t_4(3,:)=-hrc_t_4(3,:);
hrc_t_4(4,:)=-hrc_t_4(4,:);
%****************************************
% 3) Get the random data *
%****************************************
total_fft=20;
for fft_loop=1:total_fft;
for rand_data_set=1:2
rand('state', sum(100*clock));
[idat,qdat] = Pn_iq4(n,1); % -7.5 to 7.5;
%idat=ones(size(idat));
%qdat=idat;
plot(idat(1:20),'*');
grid;
title('I Data');
%pause;
iq_data=idat+qdat*1i;
plot(iq_data(100:900),'*');
grid;
title('I/Q Data Constellation');
axis([-4 4 -4 4]*2);
pause
% create a test impulse
%idat=zeros(size(idat));
%idat(200)=8;
%qdat(203)=-8;
%idat(199)=-8;
%*************************************************
% 4) Base Band Filtering & modulation *
%*************************************************
%grp_delay=taps_t/2;
offset=801+800;
nnnn=1024*4;
data_fft=10*log10(abs(fftshift(fft(kaiser(1024*4,11.)'.*...
(idat(offset:offset+1024*4-1)+qdat(offset:offset+1024*4-...
1))))));
% data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% idat(offset:offset+nnnn-1)))));
data_fft=data_fft-max(data_fft);
ff=-nnnn/2:nnnn/2-1;
ff=ff/nnnn*2*fs/2;
plot(ff,data_fft,'b');
title(' I/Q Data Spectrum Before Filtering ');
axis([0, 1e7 -85 0]);
grid;
xlabel('FREQUENCY (HZ)');
ylabel('MAGNITUDE (dB) ');
pause;
% print fig2 -dmeta;
%
% polyphase filtering with Post Filtering Carrier Recombination;
%
if_data1=4*filter(hrc_t_4(1,:),1,idat+0.5)-4*sum(hrc_t_4(1,:))*.5;
if_data2=4*filter(hrc_t_4(2,:),1,qdat+0.5)-4*sum(hrc_t_4(2,:))*.5;
if_data3=4*filter(hrc_t_4(3,:),1,idat+0.5)-4*sum(hrc_t_4(3,:))*.5;
if_data4=4*filter(hrc_t_4(4,:),1,qdat+0.5)-4*sum(hrc_t_4(4,:))*.5;
if_data=[if_data1; if_data2; if_data3; if_data4];
rf_data=reshape(if_data,1,n*4);
offset=200; % filter warm up time;
% plot spectrum;
nnnn= 1024*8;
data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
rf_data(offset:offset+nnnn-1)))));
data_fft=data_fft-max(data_fft);
ff=-nnnn/2:nnnn/2-1;
ff=ff/nnnn*2*fs/2;
plot(ff,data_fft);
title(' I/Q Data Spectrum After Filtering ');
axis([0, 1e7 -100 0]);
grid;
xlabel('FREQUENCY (HZ)');
ylabel('MAGNITUDE (dB) ');
pause;
% print fig3 -dmeta;
% axis([3e6 3.5e6 -90 0]);
% %%pause;
%************************************************
% Interpolation fs=40 MHz *
%************************************************
rf_data0=zeros(size(rf_data));
rf_data_8=[rf_data; rf_data0];
rf_data_8=reshape(rf_data_8,1,n*4*2);
fs_8=fs*2;
n_filter=160;
f=[ 0 .41 .56 1];
m=[1 1 10^(-6) 10^(-8) ];
b=fir2(n_filter, f, m);
%[H,F] = FREQZ(b,1, 1024,fs_8);
%plot(F, 20*log10(abs(H)));
%grid;
%pause
rf_data_8=filter(b,1,rf_data_8);
% plot spectrum;
% nnnn= 1024*8;
% data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% rf_data_8(offset:offset+nnnn-1)))));
% data_fft=data_fft-max(data_fft);
% ff=-nnnn/2:nnnn/2-1;
% ff=ff/nnnn*2*fs_8/2;
% plot(ff,data_fft);
% title(' I/Q Data Spectrum After Interplation x2');
% axis([0 max(ff) -100 0]);
% grid;
% xlabel('FREQUENCY (HZ)');
% ylabel('MAGNITUDE (dB) ');
% pause;
%************************************************
% Interpolation fs=80 MHz *
%************************************************
rf_data0=zeros(size(rf_data_8));
rf_data_16=[rf_data_8; rf_data0];
rf_data_16=reshape(rf_data_16,1,n*4*2*2);
fs_16=fs*2*2;
n_filter=100;
f=[ 0 .22 .77 1];
m=[1 1 10^(-10) 10^(-10) ];
b=fir2(n_filter, f, m);
%[H,F] = FREQZ(b,1, 1024,fs_16);
%plot(F, 20*log10(abs(H)));
%grid;
%pause
rf_data_16=filter(b,1,rf_data_16);
% plot spectrum;
% nnnn= 1024*8;
% data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% rf_data_16(offset:offset+nnnn-1)))));
% data_fft=data_fft-max(data_fft);
% ff=-nnnn/2:nnnn/2-1;
% ff=ff/nnnn*2*fs_16/2;
% plot(ff,data_fft);
% title(' I/Q Data Spectrum After Interplation X4');
% axis([0 max(ff) -100 0]);
% grid;
% xlabel('FREQUENCY (HZ)');
% ylabel('MAGNITUDE (dB) ');
% pause;
%************************************************
% Interpolation fs=160 MHz *
%************************************************
rf_data0=zeros(size(rf_data_16));
rf_data_32=[rf_data_16; rf_data0];
rf_data_32=reshape(rf_data_32,1,n*4*2*2*2);
fs_32=fs*2*2*2;
n_filter=60;
f=[ 0 .11 .77 1];
m=[1 1 10^(-10) 10^(-10) ];
b=fir2(n_filter, f, m);
%[H,F] = FREQZ(b,1, 1024,fs_32);
%plot(F, 20*log10(abs(H)));
%grid;
%pause
rf_data_32=filter(b,1,rf_data_32);
% plot spectrum;
% nnnn= 1024*8;
% data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% rf_data_32(offset:offset+nnnn-1)))));
% data_fft=data_fft-max(data_fft);
% ff=-nnnn/2:nnnn/2-1;
% ff=ff/nnnn*2*fs_32/2;
% plot(ff,data_fft);
% title(' I/Q Data Spectrum After Interplation X8');
% axis([0 max(ff) -100 0]);
% grid;
% xlabel('FREQUENCY (HZ)');
% ylabel('MAGNITUDE (dB) ');
% pause;
if rand_data_set == 1
%************************************************
% UP converting to 17.25MHz *
%************************************************
t=[0:n*32-1]/fs_32;
c_cos=cos(2*pi*(27.25-5-.2)*1e6*t); % carrier sine;
rf_data1=rf_data_32.*c_cos; % up converting
n_filter=351;
f=[ 0 2/8 2.2/8 3/8 1];
m=[1 1 10^(-10) 10^(-10) .001 ];
b=fir2(n_filter, f, m);
%[H,F] = FREQZ(b,1, 1024,fs_32);
%plot(F, 20*log10(abs(H)));
%grid;
%pause
rf_data1=filter(b,1,rf_data1);
% plot spectrum;
% nnnn= 1024*8;
% data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% rf_data1(offset:offset+nnnn-1)))));
% data_fft=data_fft-max(data_fft);
% ff=-nnnn/2:nnnn/2-1;
% ff=ff/nnnn*2*fs_32/2;
% plot(ff,data_fft);
% title(' RF Data Spectrum After Up Converting');
% axis([0 max(ff) -100 0]);
% grid;
% xlabel('FREQUENCY (HZ)');
% ylabel('MAGNITUDE (dB) ');
% pause;
else
%************************************************
% UP converting to 20.25MHz *
%************************************************
t=[0:n*32-1]/fs_32;
c_cos=cos(2*pi*(27.25-5+5)*1e6*t); % carrier sine;
rf_data2=rf_data_32.*c_cos; % up converting
n_filter=350;
f=[ 0 2.5/8 2.7/8 1];
m=[1 1 10^(-10) 10^(-10) ];
b=fir2(n_filter, f, m);
%[H,F] = FREQZ(b,1, 1024,fs_32);
%plot(F, 20*log10(abs(H)));
%grid;
%pause
rf_data2=filter(b,1,rf_data2);
% plot spectrum;
% nnnn= 1024*8;
% data_fft=20*log10(abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% rf_data2(offset:offset+nnnn-1)))));
% data_fft=data_fft-max(data_fft);
% ff=-nnnn/2:nnnn/2-1;
% ff=ff/nnnn*2*fs_32/2;
% plot(ff,data_fft);
% title(' RF Data2 Spectrum After Up Converting');
% axis([0 max(ff) -100 0]);
% grid;
% xlabel('FREQUENCY (HZ)');
% ylabel('MAGNITUDE (dB) ');
% pause;
end
end % of rand_data_set
rf_two_channel(fft_loop,:)=(rf_data1+rf_data2);
%rf_two_channel=rf_data2;
% plot spectrum;
nnnn= 1024*2*2*2*2;
% data_fft(fft_loop,:)=abs(fftshift(fft(kaiser(nnnn,11.)'.*...
% rf_two_channel(fft_loop, offset:offset+nnnn-1))));
end % end of fft_loop;
%temp=data_fft(1,:);
%for i=2:fft_loop
% temp=temp+data_fft(i,:);
%end
%temp=temp/fft_loop;
% data_fft_db=20*log10(temp);
% data_fft=data_fft-max(data_fft);
% ff=-nnnn/2:nnnn/2-1;
% ff=ff/nnnn*2*fs_32/2;
% plot(ff/1e6,data_fft_db);
% title(' Two-Channel 256QAM RF Spectrum');
% axis([0 max(ff)/1e6/2 -40 50]);
% grid;
% xlabel('FREQUENCY (MHZ)');
% ylabel('AMPLITUDE (dB) ');
%print fig_two_tone_QAM -dmeta;
% pause;
amp=2;
for fft_loop=1:total_fft
ctb(fft_loop,:)=0.01*((amp*rf_two_channel(fft_loop,:)).^3)+amp*rf_two_channel(fft_loop,:);
data_fft(fft_loop,:)=abs(fftshift(fft(kaiser(nnnn,11.)'.*...
ctb(fft_loop, offset:offset+nnnn-1))));
end % end of fft_loop;
temp=data_fft(1,:);
for i=2:total_fft
temp=temp+data_fft(i,:);
end
temp=temp/total_fft;
data_fft_db=20*log10(temp);
ff=-nnnn/2:nnnn/2-1;
ff=ff/nnnn*2*fs_32/2;
plot(ff/1e6,data_fft_db, 'r');
title(' Third Order Distortion');
axis([0 max(ff)/1e6/2 -30 50]);
grid;
xlabel('FREQUENCY (MHZ)');
ylabel('MAGNITUDE (dB) ');
pause;
%print fig_3rd_QAM_2 -dmeta;
%************************************************
% Add Channel Noise *
%************************************************
%************************************************
% Define Channel Model *
%************************************************
ch_delay=6-1; % 6= channel length *4 from the center tap
channel=[ -0.0 0. 0. 0. -0. 0. 0. 0. -0.0002*8*1.1 0. 0. 0. -0.0 0. 0. 0. ... % pre echo,tap index: [19 18 17 16]
0.0 0. 0. 0. -0.0 0. -0.0 0. ... % tap index: [15 14]
1 ... % center tap [13]
0.0 0. 0.0 -0.1 0. 0. 0. -0.0 0.0 0. 0. -0. 0. 0. 0. -0. ... % index: [12 11 10 9]
0. 0. 0. -0.0 0. 0. 0. -0. 0.0 0. 0. -0. 0. 0. 0. -0. ... % tap index: [8 7 6 5]
0. 0. 0. -0.+i*.0005120 0. 0. 0. -0. 0.0 0. 0. -0. 0. 0. 0. -0. ... % tap index: [4 3 2 1]
0. 0. 0.0 -0.0 0. 0. 0. -0. 0.0 0. 0. -0. 0. 0. 0. -0. ];
training=0; % if training=1 use training.
converge_i=0;
test_tap1=13-4; % start tap index of the equalizer. The index is defined in the above table [1 to 19 ]
test_tap2=13+4; % end tap index of the equalizer.
i_module=100;
end_iie=1024*4.5; % end of iie;
mu=.0015*5/2; %1/2^6; %0.03
n_test=12; % Feed Backward taps;
rf_data=filter(channel,1,rf_data);
%****************************************
% Set the SNR *
%****************************************
n_s=.2*2*1.2*(rand(size(rf_data))-.5);
%*************************************************
% 5) Demodulation *
%*************************************************
t=[0:n*4-1]/fs;
c_sin=sin(2*pi*fs*(1+.0e-3/5)/4*t+0/180*pi); % carrier sine;
c_cos=cos(2*pi*fs*(1+.0e-3/5)/4*t+0/180*pi); % carrier cosine;
rf_i_d=rf_data.*c_cos; % down converting
rf_q_d=rf_data.*c_sin; % down converting
noise_i=n_s.*c_cos; % down converting
noise_q=n_s.*c_sin; % down converting
% % plot spectrum;
% data_fft=20*log10(abs(fftshift(fft(kaiser(1024,11.)'.*...
% rf_i_d(offset:offset+1024-1)))));
?? 快捷鍵說明
復制代碼
Ctrl + C
搜索代碼
Ctrl + F
全屏模式
F11
切換主題
Ctrl + Shift + D
顯示快捷鍵
?
增大字號
Ctrl + =
減小字號
Ctrl + -