Using fixed GI, channel width and MCS

[ffrommel]

Hi.

Thank you very much for all the answers provided previously, they have been very useful to better understand Komondor.

In order to compare 802.11ax networks with legacy networks I would like to set a fixed channel width (80 MHz) and MCS (7). I have seen in the user guide that there are two parameters for that: 'basic_channel_bandwidth' and 'modulation_default', respectively. However, when I include them in the input_nodes.csv file the simulation fails: "Segmentation fault (core dumped)". Are these parameters still valid? If not, is there a way to set these two values?

Additionally I would like to select a short guard interval (GI) for testing. Is it possible to choose this? If not, what value of GI is assumed in the simulations?

Thank you very much again. Best regards,

Fabian.

[fwilhelmi]

Dear @ffrommel ,

for setting the channel width to 80 MHz, you just have to indicate the minimum and maximum channels in the nodes' input files. In your case, you will have to select 4 channels (e.g., from 0 to 3).

Regarding the MCS, I am afraid this is a deprecated parameter since we now automatically adjust the MCS based on the signal received at STAs. You can check an example of the latest nodes' input files here: https://github.com/wn-upf/Komondor/blob/master/Code/input/input_example/input_nodes.csv

Finally, regarding the GI, we have fixed several parameters for TX based on the 11ax amendment. You have all the details in https://github.com/wn-upf/Komondor/blob/master/Code/list_of_macros.h. In particular, we have fixed the duration of an OFDMA symbol based on a standard GI of 3.2 micro-seconds. If you want to change this, feel free to modify the macros in the abovementioned file "list_of_macros.h".

Best, Francesc.