Spectral efficiency in NOMA

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #14252
    Sergio Gomez
    Participant

    Dear Sir,

    I would like to calculate and plot the spectral efficiency improvements using NOMA. The simulator just gives the graphic results based in BER and Throughput.

    Would you please give some advice about how to do it?

    Thanks in advance.

    Regards,

    Sergio

    #14264
    Mariam Mussbah
    Keymaster

    Dear Sergio,

    All results (i.e. throughtput, BER, FER) are stored in the objects downlinkResults and uplinkResults.

    Best,
    Mariam

    #14272
    Kiril Kirev
    Participant

    Dear Sergio,

    if You wish to include an online calculation and plotting of the system’s SE after each simulation end, then I might also suggest a slight modification to a couple of simulator files to achieve that.

    First You’ll need to expand the properties of the ‘UserResults’ class in +Results/UserResults.m to include spectral efficiency and also to initialize it properly upon class construction. You could just copy-paste the lines for ‘Throughput’, for instance, and rename them.

    Second, for the actual parameter calculation: since the SE is closely related to the system’s throughput, You could reuse the Throughput calculation code in +Results/SimulationResults.m, lines 175-184, and divide it by the chosen system bandwidth to obtain the SE. N.B The bandwidth is not an input parameter, but is rather determined by subcarrier spacing and number of subcarriers. This info is not accessible within the ‘UserResults.m’ file, so either You’ll need to manually change the bandwidth values or further change the simulator structure in order to gain access.

    Lastly, in order to plot the SE, simply expand the code in +Results/plotResults.m in a similar copy-paste manner with the lines 123-151 and change the throughput container to the one You specified in the previous steps.

    Hope this helps!

    Best regards,
    Kiril

    #14327
    Sergio Gomez
    Participant

    Dear Kiril,
    Thank you very much for your answer.

    But now I have a question about the results.

    According to the theory, it is supposed that NOMA scheme improves the SE in the system and for FarUser, but in my simulation, when I run a no NOMA scenario vs NOMA scenario, in the second one, the SE for UE2 gets worst than in scenario1. The same results are obtained for the throughput.

    Am I doing something wrong?

    Thank you in advance.
    Regards,

    Sergio

    #14366
    Bashar Tahir
    Participant

    Dear Sergio,

    The implementation of NOMA in our simulator is based on the 3GPP MUST scheme in which the farUE is limited to QPSK modulation. The scheme is to be used such that a strong cell-center UE is coupled with a weak cell-edge UE that cannot transmit with high rate anyway to begin with.

    If you switch to OMA, and let the transmit power of the BS increases, then sure, at some point the farUE will no longer be weak (since it now has very high SNR), and then it would comfortably switch to high modulation orders and code rates, hence, resulting in a higher throughput.

    Best,
    Bashar

    • This reply was modified 11 months, 1 week ago by Bashar Tahir.
    • This reply was modified 11 months, 1 week ago by Bashar Tahir.
    #15249
    Farhan
    Participant

    Dear Bashar,

    For massive MIMO scenario, if we are using multiuserMode.Downlink={‘ZF-MUMIMO’} and the schedule for the superimposed users below, while playin it says the schedule in the scenario file is ignored and all users assigned to the base station are scheduled to transmit/receive on all available subcarrriers. Is it reasonable to say that we are still using NOMA (3GPP MUST technique) with these setting? Or when using ZF or MRT we are using OMA?

    scStr.schedule.fixedScheduleDL{1} = [‘UE1:72,UE2:UE1,UE3:UE1,UE4:UE1’];
    scStr.schedule.fixedScheduleDL{2} = [‘UE5:72,UE6:UE5,UE7:UE5,UE8:UE5’];
    scStr.schedule.fixedScheduleUL{1} = [‘UE1:72,UE2:UE1,UE3:UE1,UE4:UE1’];
    scStr.schedule.fixedScheduleUL{2} = [‘UE5:72,UE6:UE5,UE7:UE5,UE8:UE5’];

    #15264
    Bashar Tahir
    Participant

    Dear Farhan,

    In principle, it is a non-orthogonal access, since the users transmit over the same resources; however, such a type of access falls under space-division multiple access (SDMA). So no, it is not 3GPP MUST and nor conventional NOMA in a direct sense. Moreoever, in order to benefit from NOMA, you need power-domain multi-user detection, which can be based on a simple interfernce cancellation or through a more-complex joint maximum-likeihood search, which are both not enabled anyway when MUMIMO modes are used.

    In the current release of the simulator, only downlink NOMA is supported, and it is enabled by setting multiuserMode.Downlink to MUST. In this case, joint maximum-likelihood detection over the composite constellation of the near- and far-UEs is performed at the near-UE, while the far-UE detects its signal directly ignoring the interference by near-UE.

    Best,
    Bashar

Viewing 7 posts - 1 through 7 (of 7 total)
  • You must be logged in to reply to this topic.