Reproducing the results in the paper

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
    Posts
  • #5712
    Omar Sayan
    Participant

    Hello!
    I’m new to LTE simulator and just starting to learn to use it.

    I am using the LTE Link Level simulator version 1.3 and I’m trying to recreate the results that are provided in the documentation that comes with the simulator. I am running LTEA_sim_batch_quick_test_SUMIMO.m for the 8×1 case with 1000 subframes and SNR vector = [0:3:21].

    But the result resembles the figure in the paper only remotely: in the paper, throughput is near 2.5 Mbit/s at SNR = 5. However in my results the throughput is 0 at SNR = 5. Also the saturation happens at nearly 3.5 Mbit/s, while in the paper it happens at 5 MBit/s. my result

    Is this normal? I expected to be able to produce the exact result as in the provided paper, but that’s not the case.

    Thank you!

    • This topic was modified 5 years, 1 month ago by Omar Sayan.
    #5714
    Omar Sayan
    Participant

    Image and video hosting by TinyPic“>

    #5725
    Stefan Schwarz
    Participant

    Dear Omar,

    your results look as if you are simulating with a fixed CQI, since the delay is zero for a long time and finally saturates below the maximum. Probably you have deactivated the CQI feedback (LTE_params.UE_config.feedback = false or LTE_params.UE_config.CQI_fb = false in LTE_load_paramters)?

    Best,
    Stefan.

    #5726
    Stefan Schwarz
    Participant

    b

    #5727
    Stefan Schwarz
    Participant
    #5729
    Omar Sayan
    Participant

    Thank you for your reply, Dr. Schwartz!

    Strangely, both parameters are set to true.
    Lte Parameters snapshot

    I’d like to also note that when I load the file 8x1_results.mat in the examples/LTEA SUMIMO quick test results folder and plot the results, it gives the same plot as in the paper (which is good). But the fields in LTE parameters are missing in the .mat file that I newly generated in my simulation run. Not sure if this is important though.

    – My simulation parameters fields

    – parameter fields in the provided results in the ‘examples’ folder

    Thanks!

    #5733
    Omar Sayan
    Participant

    I am still obtaining this inaccurate result. Like I said in my previous post both parameters are set to true. Also, I am wondering how does the number of subframes affect the simulation? I tried to set the feedback and cqi to false, but that resulted in an error.

    Error using LTE_feedback (line 110)
    Attempted to access CQI(1,1); index out of bounds because size(CQI)=[0,0].

    I would greatly appreciate your help.

    Thanks

    #5734
    Stefan Schwarz
    Participant

    Okay, so feedback seems to be activated.

    Again, are we talking here about version 3 of the LTE-A simulator? Because I cannot reproduce your issue when I run it on my computer with this version. Did you change any parameter?

    Best,
    Stefan.

    #5735
    Omar Sayan
    Participant

    Yes, this is version 3.
    I haven’t changed any parameters, I can send to your email all of my workspace variables saved in a .mat file just to confirm all parameters are correct.
    Once again I’m simply trying to recreate the results in the provided paper and I’m following the steps in it (running LTEA_sim_batch_quick_test_SUMIMO.m).

    Regards,
    Omar

    #5736
    Stefan Schwarz
    Participant

    Yes that might be helpful. Please save the variable “LTE_params” after the call of “LTE_load_parameters_generate_elements” in “LTE_sim_main_process_SNR_vector.m” and send it to sschwarz@nt.tuwien.ac.at

    Thanks,
    Stefan.

Viewing 10 posts - 1 through 10 (of 10 total)
  • You must be logged in to reply to this topic.