I want to implement my own mode selection method in OMNeT++ 5.0 using the simulte-0.9.1 framework.
According to the demos and the paper from the authors, mode selection should generally be enabled liked this in the .ini:
*.eNodeB.nic.phy.enableD2DCqiReporting = true
*.eNodeB.nic.d2dModeSelection = true
*.eNodeB.nic.d2dModeSelectionType="D2DModeSelectionBestCqi"
The BestCQI mode selection method is already implemented and would serve as a nice example on how to implement new methods. Unfortunately OMNeT++ complains that all three are unused parameters. I simulated something for a short while and had a look at the logs. In simulte/src/stack/d2dModeSelection/d2dModeSelectionBestCqi/D2DModeSelectionBestCqi.cc the very first line in the doModeSelection() function it says
EV << NOW << " D2DModeSelectionBestCqi::doModeSelection - Running Mode Selection algorithm..." << endl;
but no such output comes up when I run the simulation, so it doesn't seem to be working.
It appears that these parameters actually work, even though they're marked unused. I simulated two scenarios: one with mode selection off and one with on and had a look at different throughput scalars.
Throughput with mode selection off
Throughput with mode selection on
The eNB sees a lot of traffic when these parameters are present. I don't see how the traffic could be as large as it is in my simulation, but at least it seems to have an effect.
