Hi Bruno and all, talking about distance tests - if you did that in 802.11a or g, there is problem with ack timeouts and slottimes - in madwifi it has to be adjusted manually for longer distances, and from my experience with ath5k there are problems with acks even when distance is 0,5m, so maybe txpower is not the problem here, or not the only problem. I will do some more tests and let you know, from what i already made :
madwifi as ap, ath5k as sta, ff off, bursting off, 802.11a , various tests with fixed rate to 18 ,24,36,48 and some with rate auto on both sides. Distance about 30cms. Used cards - CM10 in minipci2pci holder. RSSI on madwifi side of about 40db, jumping to 60db occasionaly. Maximum throughtput has been on rate auto n both sides, in the best case it has been 24,3Mb/s in iperf TCP test. Madwifi2madwifi has been able to do about 33Mb/s in average in same conditions, and about 54Mb/s on channel 152 in turboa mode. Main difference between madwifi2madwifi and madwifi2ath5k has been, that, when watchng stats on madwifi side of link - athstats -i wifi0 1 - in case of madwifi2ath5k i have seen 2x more input frames that output frames. On madwifi2madwifi both sides has been about the same - with small fluctuations. From that fact i think there could be some problem with acks. Will study it a bit - recently haven't had much time for this - and report more. Pavel _______________________________________________ ath5k-devel mailing list ath5k-devel@lists.ath5k.org https://lists.ath5k.org/mailman/listinfo/ath5k-devel
