Mismatch in HE (802.11ax) SU PPDU duration (see figure 27-8 of IEEE Std 802.11ax-2021) between NS3 and IEEE Std 802.11ax-2021

34 views
Skip to first unread message

Faisal Darbari

unread,
Feb 18, 2025, 4:45:16 AM2/18/25
to ns-3-users
Hi There, 
I'm new to NS3. I've noticed a mismatch in HE SU PPDU duration calculate in NS3. Am I missing anything?

PPDU duration of HE SU PPDU (ref: figure 27-8 of IEEE Std 802.11ax-2021)

  • Scenario: (HE-SU, GI = 3.2, MCS-7, APEPLength = 1536 bytes). This calculation is based on the standard
    • Preamble (L-STF + L-LTF + L-SIG) = 8 µsec + 8 µsec + 4 µsec = 20 µsec
    • Header (RL_SIG + HE-SIG-A + HE-STF + HE-LTF) = 4 µsec + 8 µsec + 4 µsec + 16* X µsec  = 16 + 16X µsec (Where X=1,2, and 4 for NSS1, NSS2, and NSS3/4 respectively, 16 is the HE-LTF symbol duration for GI=3.2)
    • Preamble+Header = 20 µsec + 16 + 16X µsec = 52 µsec for NSS1
    • Payload duration = 176 µsec
    • PPDU duration = Preamble+Header+Payload duration = 52 µsec + 176 µsec = 228 µsec

PPDU duration of HE SU PPDU (in NS3) for the same scenario above

    • Preamble (L-STF + L-LTF + L-SIG) = 8 µsec + 8 µsec + 4 µsec = 20 µsec
    • Header (RL_SIG + HE-SIG-A + HE-STF + HE-LTF) = 4 µsec + 8 µsec + 4 µsec + 8* X = 16 + 8X µsec,  for X=1 (Where X=1,2, and 4 for NSS1, NSS2, and NSS3/4 respectively, 8 is the HE-LTF symbol duration for GI=3.2 instead of 16)
    • Preamble+Header = 20 µsec + 16 + 8X µsec = 44 µsec for NSS1 (i.e X=1)
    • Payload duration = 176 µsec
    • PPDU duration = Preamble+Header+Payload duration = 44 µsec + 176 µsec = 220 µsec

The HE-LTF symbol duration of 8 µsec for Guard Interval 3.2 seems incorrect in NS3. This should be 16 µsec (ref: Table 27-12 of IEEE Std 802.11ax-2021). Is this correct?

Thanks,

Faisal

Tom Henderson

unread,
Feb 19, 2025, 9:54:47 PM2/19/25
to ns-3-...@googlegroups.com
Thanks for reporting this; one of our wifi module maintainers will have to respond.  So that we do not lose track of this, I created this issue in the tracker: https://gitlab.com/nsnam/ns-3-dev/-/issues/1188

- Tom
Reply all
Reply to author
Forward
0 new messages