Date: Sat, 22 Sep 2012 12:36:11 -0700 From: Adrian Chadd <adrian.chadd@gmail.com> To: Johann Hugo <jhugo@meraka.csir.co.za> Cc: freebsd-wireless@freebsd.org Subject: Re: Performance of AR9287 Message-ID: <CAJ-VmokwirNTy=zCOjxEq_jUF6ra6fzp4O2OE9%2Bb4v%2BGC5CkCQ@mail.gmail.com> In-Reply-To: <CAJ-VmonEhSJ_%2BJqPS7w5G03SkSVjQoLBviBE=CrgxkoUj-53XA@mail.gmail.com> References: <201209202357.06358.dantavious313@gmail.com> <201209221020.16702.jhugo@meraka.csir.co.za> <CAJ-VmomWU8qWVgRws7=ewVX%2BxZZ-2L-0KBB4d=q1Uijsfq5u-g@mail.gmail.com> <201209221104.43360.jhugo@meraka.csir.co.za> <CAJ-VmonEhSJ_%2BJqPS7w5G03SkSVjQoLBviBE=CrgxkoUj-53XA@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
.. and yes, what I'd really like to do is find someone who wants to look after "long distance links" with atheros hardware (and has the kit to do so!) and help me fix up the coverage class code to work in a more generic fashion, for all chips. I can tell people what to do if they'd like to hack on this (specifically look at ah_coverageClass, and ar5212SetCoverageClass()) and I can help them with the math, I just don't have the time to make it all work right. :-) Adrian On 22 September 2012 12:33, Adrian Chadd <adrian.chadd@gmail.com> wrote: > Ah. athctrl.sh ? > > > > > adrian > > On 22 September 2012 02:04, Johann Hugo <jhugo@meraka.csir.co.za> wrote: >> I'm still using a utility from Sam. >> >> >> >> athctrl -d 8000 >> >> Setup IFS parameters on interface ath0 for 8000 meter p-2-p link >> >> dev.ath.0.slottime=36 >> >> dev.ath.0.acktimeout=75 >> >> dev.ath.0.ctstimeout=75 >> >> >> >> Are these settings still good ? >> >> >> >> Are there any way do detect these automatically ? >> >> >> >> Johann >> >> >> >> On Saturday 22 September 2012 10:24:43 Adrian Chadd wrote: >> >>> I assume you've tweaked the ack, slot, rtsctstimeout? >> >>> >> >>> The 11n rate control code is a little more jumpy. I've fixed that in >> >>> recent checkins. It's not optimal but now it's not as jumpy as before. >> >>> >> >>> Update to -HEAD, then use 'athratestats' on the peer MAC address. It's >> >>> quite useful to see what's actually going on. :-) >> >>> >> >>> >> >>> >> >>> Adrian >> >>> >> >>> On 22 September 2012 01:20, Johann Hugo <jhugo@meraka.csir.co.za> wrote: >> >>> > On Saturday 22 September 2012 02:05:23 Derrick Edwards wrote: >> >>> >> Thanks again for looking into this. >> >>> >> >> >>> >> My main concern was the range of network performance that I was getting >> >>> >> >> >>> >> when the laptop was right in front of the access point. As you can see >> >>> >> >> >>> >> bandwith varied greatly. >> >>> > >> >>> > I don't know if it's related, but my RSSI and RATE jumps around quite a >> >>> > lot while transfering a large file. >> >>> > >> >>> > >> >>> > >> >>> > Output from loop doing "ifconfig wlan0 list sta" while transfering a big >> >>> > file. >> >>> > >> >>> > >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 86M 16.5 0 842 9440 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 43M 12.5 0 1063 14784 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 43M 15.5 0 1362 21088 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 86M 15.5 0 1627 27488 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 86M 13.5 0 2150 38976 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 86M 11.5 0 2289 42384 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 29M 24.5 0 3075 60016 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 86M 28.5 0 3369 1600 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 86M 14.0 0 3839 11744 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 29M 13.5 0 87 19936 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 58M 11.5 0 403 27440 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 58M 12.5 0 758 35200 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 58M 11.0 0 1054 42320 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 58M 23.5 0 1249 46512 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 29M 11.0 0 1381 49312 E AQHTR HTCAP WME >> >>> > >> >>> > ADDR AID CHAN RATE RSSI IDLE TXSEQ RXSEQ CAPS FLAG >> >>> > >> >>> > 00:0c:42:6c:58:29 1 128 58M 11.5 0 1584 54224 E AQHTR HTCAP WME >> >>> > >> >>> > >> >>> > >> >>> > System info = hybrid >> >>> > >> >>> > HW = gateworks ARM >> >>> > >> >>> > OS = FreeBSD-9 with ath code from current >> >>> > >> >>> > Wifi = RouterBOARD R52nM = AR9220 >> >>> > >> >>> > Distance = 8km >> >>> > >> >>> > >> >>> > >> >>> > B.t.w if I use the same HW with FreeBSD-9 in 11a mode, then RSSI does >>> > not >> >>> > jump around and I get about the same throuput as with N. >> >>> > >> >>> > >> >>> > >> >>> > I don't have any athstats output, still working on it. >> >>> > >> >>> > >> >>> > >> >>> > Johann >> >> r
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAJ-VmokwirNTy=zCOjxEq_jUF6ra6fzp4O2OE9%2Bb4v%2BGC5CkCQ>