Not getting throughput as in MCS table

36 views
Skip to first unread message

Pedro Turbo

unread,
May 25, 2023, 2:16:52 AM5/25/23
to ns-3-users
Hi all,

I'm using the wifi-he-network example.
Here's how I run it:
./ns3 run "wifi-he-network --frequency=6 --distance=0.01 --udp=1 --mcs=11 --payloadSize=2000 --phyModel=Yans"

This is the output:
MCS value               Channel width           GI                      Throughput
11                      20 MHz                  3200 ns                 111.677 Mbit/s
11                      20 MHz                  1600 ns                 111.645 Mbit/s
11                      20 MHz                  800 ns                  130.334 Mbit/s
11                      40 MHz                  3200 ns                 213.531 Mbit/s
11                      40 MHz                  1600 ns                 213.555 Mbit/s
11                      40 MHz                  800 ns                  247.475 Mbit/s
11                      80 MHz                  3200 ns                 408.434 Mbit/s
11                      80 MHz                  1600 ns                 408.192 Mbit/s
11                      80 MHz                  800 ns                  466.24 Mbit/s
11                      160 MHz                 3200 ns                 691.84 Mbit/s
11                      160 MHz                 1600 ns                 691.779 Mbit/s
11                      160 MHz                 800 ns                  775.083 Mbit/s

According to the MCS table (https://mcsindex.com/), for 160MHz, mcs 11, 800 usecs GI for example, I should be getting 1201 mbits/sec, but getting only 775.

I know UDP and IP headers have fixed overheads, but it doesn't explain the whole difference.
What am I missing?

Tommaso Pecorella

unread,
May 25, 2023, 5:10:05 AM5/25/23
to ns-3-users
Hi,

it might be a bug (you never know), but don't underestimate the WiFi overhead. The table you used reports the "raw" bitrate, but WiFi can't use it all - it have to add pretty long headers and guard intervals between packets...

Reply all
Reply to author
Forward
0 new messages