Date: Tue, 4 Oct 2005 18:43:29 -0500 From: Sam Pierson <samuel.pierson@gmail.com> To: freebsd-mobile@freebsd.org Subject: Different ACK response times in ath? Message-ID: <d9204e4c0510041643l6d9d0b4as5e86c819d73c13cd@mail.gmail.com>
next in thread | raw e-mail | index | archive | help
Hi all,
Recently I had to do some measurements and came up with some interesting
results with the ath driver. I would take two machines associated to an AP
and ping the second with the first. This produces four data frames:
A --> AP : echo request
AP --> B: echo request
B --> AP: echo response
AP -> A: echo response
Each of these directed data frames results in an ACK packet being
generated by the destination. I monitor all traffic using a passive
third laptop, which also records elapsed time between frames
(specifically between data frames and ACK packets).
Under channel 11, which has a lot of traffic, I had these results:
(times refer to elapsed time between data and ACK frames)
Using an IBM TP T42:
D-link DWL-G630: ~248us
Netgear WG511T ~210us
Using IBM TP X23:
D-link: ~208us
WG511T: ~210us
internal prism 2.5: ~256us
Apple Airport:
~250us
Why would these values be different?
Additionally, on channel 6 (with little traffic), when using
one of the machines in hostAP mode:
X24: netgear: 114us
dlink: 109us
airport: ~154us
Shouldn't ACK packets be transmitted a constant delay after
directed data frames have been received? Isn't this the way
the 802.11 standard is defined? Why is the slower computer
responding faster than the faster computers when the TPs
are running the same drivers (and versions)?
--
Sam
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?d9204e4c0510041643l6d9d0b4as5e86c819d73c13cd>