Re: [bmwg] an ID for BMWG to consider

On Fri, 29 Oct 2004, Kevin Dubray wrote:
> The implication on IP devices is an important component to consider.
> Anyone care to list those implications for the less than
> 802.11 savvy amongst us?  (It appears the proposed document hopes
> that RFC1242 et al. makes that binding.)

Here are just some quick examples off the top of my head. I'm sure there 
are others:

1. Throughput -- The air is a very lossy medium, and thus zero-loss 
measurements such as throughput aren't meaningful. We should not abuse the 
1242 definition by attempting to label as "thoughput" measurements that 
involve loss.

2. Delay, jitter, and other time-related metrics -- The 802.11 protocols 
acknowledge every single frame at L2 and also send significant amounts of 
control traffic.

While it's possible (though uncommon) to measure delay of individual 
packets or instantaneous jitter from two packets, those probably won't be 
meaningful indicators of how long a DUT/SUT holds up application traffic. 
There are probably more meaningful, WLAN-specific goalposts to be defined.

3. 802.11-specific stuff -- There are some WLAN events, such as roaming 
time, that existing bmwg methodologies do not describe.

dn

_______________________________________________
bmwg mailing list
bmwg@ietf.org
https://www1.ietf.org/mailman/listinfo/bmwg