A question about ploting results

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
    Posts
  • #14157
    Cold Heart
    Participant

    Dear sir,
    An error result will be outputed, when setting nFrames = 1 in massiveMIMO.m in some situations.
    For example, setting sweepParam = {‘simulation.pathloss’}, sweepValue = [100, 90] and plotOverSNR = true, the xValues in plotResults.m will be the mean SNR calculated by set pathloss in massiveMIMO.m.

    The reason is function mean(a) is equal to mean(a, 1), when size(a, 1) > 1; mean(a) is equal tp mean(a, 2), when size(a, 1) = 1.
    When setting nFrames = 1, size(results.userResults{iUE}.SNR, 1) = 1, which leads to the problem.

    Modifying “xValues = mean(results.userResults{iUE}.SNR);” to “xValues = mean(results.userResults{iUE}.SNR, 1);” is suggested in plotValues.m.
    Modifying “obj.userResults{iUE}.xxx.mean = mean(obj.userResults{iUE}.xxx.values);” to “obj.userResults{iUE}.xxx.mean = mean(obj.userResults{iUE}.xxx.values, 1);” is suggested in SimulationResults.m, where xxx could be BERCoded, BERUncoded, FER, throughput, channelMSE and so on.

    Best Regards,
    Xin Liang

    #14271
    Kiril Kirev
    Participant

    Dear Mr. Liang,

    thank You for Your input! Yes, indeed without unambiguous dimension specification in the mean function, this error will be produced each time the parameter scStr.simulation.nFrames is set equal to 1 in any scenario file.
    Since results are calculated as an average over multiple (many) random channel realizations (the Monte Carlo method), averaging over large number of realisations (the aforementioned parameter scStr.simulation.nFrames) leads to higher statistical significance of the results. Thus for many simulation setups, setting nFrames to 1 does not make much sense.
    Of course, if Your specific simulation setup does not include random realizations and the plots in +Results/plotValues.m are still useful for Your purposes, please feel free to alter the code. Otherwise, using multiple (many) frames to average over is recommended.

    Best Regards,
    Kiril

    #14369
    Roaa Ali
    Participant

    Dear sir
    I use vienna 5G link level simulator version 1.2 and matlab R 2018b. When implement program and choosing scenario ‘NOMA’. plot BER begin decrease but change plot to increasing
    What is cause that?
    Thanks
    Best regards,
    ROAA

    #14374
    Bashar Tahir
    Participant

    Can you please clarify what the problem is? It is not clear to us what you mean.

Viewing 4 posts - 1 through 4 (of 4 total)
  • You must be logged in to reply to this topic.