User Tools

Site Tools


group:status:mimo

HomeResearchMembersPublicationHonorCoursesServicesNewsOpeningsGroup

MIMO Project

Personnel

  • Hussein Saad (Coordinator)
  • Zhou Zhou
  • Rubayet Shafin
  • Shashank Jere

Tasks and Deadlines

  • SVD and data sheet (due: 09/15/19)
  • Annual report (due: 09/17/19)
  • Actions needs to be done by 09/23/19
  • Implementation of initial MU-MIMO scheduling algorithm for ZF-MMSE. Due 10/03/2019
    • In the SVD, for the link level simulator, you call out two alternate ways of running the simulation, either without channel encoding or with. Rather than give the user a choice about which of these scenarios to execute, I would like these two scenarios called out separately in the SVD as unique scenarios, each with a visual indicator of expected outputs when it is executed. I have attached an example from the VT IM working group as an example of what I am looking for.
      • Action #1: Zhou, Shashank: Please update the SVD file to indicate there are two unique scenarios, and give visual outputs indicating what each of those scenarios will render.
    • When I download the execute the link level simulations, both for the channel encoding and without the channel encoding, the plots don't match the plots in the SVD. For one thing, the grid is off for some reason. I went ahead and turned the grid on, but beyond that, I encourage you to go out to the dev branch of the repo, pull down the files, and run them - and see if the plots align. Mine don't. Just for reference, I've included plots for the link level LDPC as-is, without modification, from when I ran on my local machine.
      • Action #2: Zhou, Shashank Please fix the simulation files for the link level simulator (both cases) so that the simulation outputs and the SVD match.
    • Rubayet I'm in the process right now of running the main_SINR_new file to verify that the output plot matches that in the SVD, but it takes a long time on my machine. I'm asking you to verify this as well and, if necessary, make any modifications to the Matlab code. (Something I find really annoying is when the grid lines don't automatically turn on.)
      • Action #3: Rubayet, Shashank Please verify the simulation files for the system simulator match the SVD plots, and if necessary, update the m-files (or SVD, as appropriate).
    • Only one of the three Matlab files submitted to generate output for the SVD have CI/CD tests enabled. There are three unique actions I need you to take on this.
      • Action #4: Shashank, Zhou For the link simulator without channel encoding, when I ran it at one point, the CI/CD tests FAIL. Need this test criterion fixed or relaxed so that a meaningful pass/fail criterion is established.
      • Action #5: Shashank, Zhou For the link simulator without channel encoding, I need you to add a CI/CD test with a pass and fail criterion.
      • Action #6: Zhou, Rubayet For the System simulator, I need you to add a CI/CD test with a pass and fail criterion.
    • The dev branch has been modified across the board to include a standard RRPUCS file header for submission to the customer. I need you to keep these intact - if you modify anything, only modify the comment field describing the function of the module. Do not modify anything above the function and file name part of the header.
      • Action #7: Zhou, Rubayet, Shashank Whatever files you modify, please only change the dev branch. If you must add files, please identify all of those files to me in a separate list so that I can include the appropriate file headers for the program. Please do not insert the program file headers yourself by copying and pasting from another file. We have an automated script that does that work. Above all, please do not modify the master repo. You will need to do a git pull to get your copy of the repo to equal what is out on the git server.
      • Action #8: 1) Proof read the documents, 2) make sure the fonts are “Times New Roman”, and 3) all the generated simulation results match well with the documents. (Zhou, Rubayet, Shashank)
    • White paper for future directions including machine learning for channel estimation and beamforming submitted to GIRD. Feedback was received and responses have been sent out. (Due 10/21/2019)
      • Six months plan for 5G and machine leanring direction to be sent out by 10/31/2019
  • Future research directions: A literature survey was sent to GIRD about ML-based channel estimation. A matlab code for data generation as well as a pytorch code for the denoising idea is under preparation. Due 11/21/2019

Execution

  • 12/16/19: Meeting with GIRD. Presented results of Link-level simulator with OFDM integrated and initial Denoising results from NN Pytorch implementation. Weekly Meeting Minutes
  • 11/25/19: Meeting with GIRD. Discussed requirements for code drop on Dec 10th. Weekly Meeting Minutes
  • 11/18/19: Discussion with GIRD on proposed Neural Network-based MIMO-OFDM Channel Estimation. Weekly Meeting Minutes.
  • 11/11/19: Meeting with GIRD, submitted Literature Review on NN-based Channel Estimation. GIRD provided initial OFDM packet structure. Literature Review OFDM Functional Description
  • 11/04/19: Discussion of 6-month plan with GIRD. Six-month planMinutes and AIs.
  • 10/03/19: Implemented CBUS scheduling algorithm for ZF-MMSE precoding with perfect CSI availability. Prepared the weekly summary ppt and bi-weekly report.
  • 09/20/19: Internal meeting (Hussein, Rubayet, Shashank, John)
  • 09/19/19: Weekly report to GIRD is submitted.
  • 09/17/19: Annual report to GIRD is prepared.
  • 09/15/19: Matlab code for link level and system level simulators are updated.
  • 09/15/19: Software Version Description (SVD) and Data Sheet are completed.
group/status/mimo.txt · Last modified: 2019/12/16 16:08 by shashankjere