Re: [DNA-BOF] minimized DAD in WLAN

["Nick 'Sharkey' Moore" <sharkey@zoic.org>]

On 2003-09-01, Soohong Daniel Park wrote:
> 
> I've written minimized DAD with new timer and parameters
> instead of current DAD.

Okay, so we're back to just reducing the timers again?
We've done this a few times in mobileip.

The problem is that you have to decide how far you reduce them.
50ms may seem a fair number, but:

* >10ms is a long time considering our non-predictive L2 handovers
	were completing in <50ms[*].

* <100ms is a short time given that a collision or dropout
	at L2 may delay the NA, and that 'sleeping' nodes may
	take some time to respond.

This makes it difficult to select an appropriate delay.

cheers,

-----Nick
[*] the usual caveats about testbed situations apply.
-- 
Nick 'Sharkey' Moore  <sharkey@zoic.org>  <http://zoic.org/sharkey>
"I'm just going to get my feet wet until I drown." -- Ani DiFranco