Date: Mon, 27 Oct 2014 02:02:42 +0000 From: bugzilla-noreply@freebsd.org To: freebsd-wireless@FreeBSD.org Subject: [Bug 194620] New: [ath] [net80211] take ampdu length and error rate into account when select ampdu length Message-ID: <bug-194620-21060@https.bugs.freebsd.org/bugzilla/>
next in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=194620 Bug ID: 194620 Summary: [ath] [net80211] take ampdu length and error rate into account when select ampdu length Product: Base System Version: 11.0-CURRENT Hardware: Any OS: Any Status: Needs Triage Severity: Affects Only Me Priority: --- Component: wireless Assignee: freebsd-wireless@FreeBSD.org Reporter: adrian@freebsd.org Right now the A-MPDU error logic in ath_rate_sample only looks at the error rate, not at where the errors are. eg: http://www.mwnl.snu.ac.kr/~schoi/publication/Conferences/14-CONEXT-BYEON.pdf so for longer frames, things like changing channel conditions, clock drift/phase drift and such will cause the BER for the later subframes to be greater than the earlier frames. The atheros NICs do channel estimate at the beginning of each frame. So, let's add a bit of logic to check the RX AMPDU BER pattern and if we see the later sub-frames with a higher BER than the earlier subframes, start shortening the AMPDU. -- You are receiving this mail because: You are the assignee for the bug.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-194620-21060>