why CSI magnitude values are more consistent compared to the CSI amplitudes?

[esoltana]

Hi, does any body know why the CSI magnitude/SNR values are more consistent compared to the raw scaled CSI amplitudes over sequential packets for all the subcarriers? Basically I compared the two graphs using plot(db(abs(squeeze(csi).'))) and plot(db(real(squeeze(csi).'))) and the second graph has much more variation over time.

[zainalabidenakhtar]

@soltanaghaei can you please share any code to plot with respect to time because i also want to plot CSI with time instead of subcarrier index but not getting correct result may be i am doing some mistake. thanks in advance

[azizshahics]

@zainalabidenakhtar Please consider following code for time domain analysis

csi_trace=read_bf_file('june\p1b.dat'); %%% reads the CSI file p1b.dat csi_entry = csi_trace{1} %%% trace the 1st packet of CSI data [each packet %%% contains SNR vs 30 subcarrier information] csi=get_scaled_csi(csi_entry) db(get_eff_SNRs(csi),'pow')

ya=(db(abs(squeeze(csi).')));         %%% ya represents SNR vs 30 subcarreirs for 3 Rx Antennas

  antenna1a=ya(:,1);                        %% Rx antenna A
antenna2a=ya(:,2);                              %% RX antenna B
antenna3a=ya(:,3);                              %% RX antenna C

%% Time Domain [ X axis extracting code ] i=1:30; n=30; % number of subcarriers fop=2010^6; % 2.43 GHz operating frequency of wifi router t=[(fop/2)/n]^-1; % formula taken from mathworks di=t(i-1);

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

power_time_domain=20*log10(abs(ifft(antenna2a))); %% Y axis power plot (same as SNR) plot(di,power_time_domain)

[zainalabidenakhtar]

@azizshahics thanks a lot for your favour

[zainalabidenakhtar]

how the value is 'i' is selected? is this same as subcarriers?

[zainalabidenakhtar]

when I run this code I actually get this error:

Subscript indices must either be real positive integers or logicals.

Error in time_domain_plot (line 32) di=t(i-1);

[azizshahics]

yes its the same as sub-carrier number.

[ogutiann]

@zainalabidenakhtar where you able to solve this error?