Re: [Madwifi-devel] disable cca but low pps at the receiver
Status: Beta
Brought to you by:
otaku
From: Vishal S. <vis...@gm...> - 2010-06-26 17:19:03
|
what value are you setting for disable_cca?? vishal On Sat, Jun 26, 2010 at 10:23 PM, zbdou <dz...@fo...> wrote: > I tried, but no improvements. > > I changed the channel to 3, which has less external interference. > With cca enabled, the maximum throughput is 4500pps (54Mbps, 1000-byte > broadcast frames). > but after disable cca at the sender, > if we set the sending interval of pktgen to tx 4500pps, > the receiver could only get throughput of 400pps, so strange! > > I think there is something wrong with the patch. > > > > > > > try increasing the acktimeout using > > sysctl -w dev.wifi0.acktimeout=30 > > play around with couple of values and see what happens. from what i have > observed value of around 25 us works .. with slightly higher value, the cca > is not disabled :) (which i find strange) > > On Sat, Jun 26, 2010 at 9:49 PM, dead cow <dz...@fo...> wrote: > >> Hi, vishal, >> >> I send boradcast frames using pktgen. >> The ubiquity XR2 cards are used for this test. >> >> >> zbdou. >> >> >> >> >> Have you disabled the ack also or ack is enabled?? and what hardware are >> you using?? >> >> for ar5212 based ubiquiti and mikrotik cards what i observed is if i do >> large file transfer of 20-30 mb using scp, then the transfer stall midway. >> so am doubtful if the patch works accurately. >> >> vishal >> >> >> On Sat, Jun 26, 2010 at 9:34 PM, dead cow <dz...@fo...> wrote: >> >> Hi, all >> I have disabled cca/ csma,ca based on: >> https://systems.cs.colorado.edu/projects/carp/browser/trunk/patches/noCCA.patch >> It really works. >> But the problom is >> when the sender sends at the full speed at 802.11g, fixed rate 54Mbps, >> 1000-byte frame, (around 5500 packet per frame), >> the receiver could receive almost no frames. >> >> This situation is substantially mitigated when the sender transmits at a >> lower speed (e.g. 2600pps), >> at the receiver, we got 2300pps. >> >> Does anyone know the reason about this strange problem? >> >> Thanks! >> >> >> 2010-06-26 >> >> >> >> dead cow >> >> >> ------------------------------------------------------------------------------ >> This SF.net email is sponsored by Sprint >> What will you do first with EVO, the first 4G phone? >> Visit sprint.com/first -- http://p.sf.net/sfu/sprint-com-first >> _______________________________________________ >> Madwifi-devel mailing list >> Mad...@li... >> https://lists.sourceforge.net/lists/listinfo/madwifi-devel >> >> >> > |