Re: [DNA] Scope of solutions discussion document

[Greg Daley <greg.daley@eng.monash.edu.au>]

Erik Nordmark wrote:
> Greg Daley wrote:
> 
>> Here's the 30 hosts incidence plot for the delay after RS that an RA
>> was received using one millisecond buckets. The incidences were
>> summed over the simulation runs. Please be aware that the values are for
>> the middle of the millisecond range (so it works for box plots).
>>
>> The actual delay for beginning to probe until the reception of the RA
>> was much more spread out. I can send similar data to individuals
>> in order to get an impression, but I'd prefer not to send larger
>> attachments.
> 
> 
> In these tests, when did the hosts send the RS? Was it done after 802.11 
> had been associated (and authenticated if .1x is used), or earlier?

There was no 802.1x, just relying on 802.11/WEP.

RS was immediate on reception of L2 UP indication. (on
completion of association/reassociation). No induced delay
was created for cryptographic operations.

The RSs were queued to the MAC without other packets being
in the queue (the moving nodes weren't transmitting ...
a bit of a foible, but one which can be effected by
having a priority queue).

RAs were added to the FIFO AP queue.

> If we take the approach that the L2 hint to DNA is delivered after 
> association (and authentication, when used), then this would tend to 
> spread out the RSes a bit, which might be sufficient.

I'd guess that the use of 802.1x or similar would provide more 
serialization, but would increase the MAC contention, and lengthen
the time to exchange upper layer packets.

The principle may still hold though.

It looks like I won't have time before 62 to generate simulations of
effects on the MAC for existing users, since I've maxed out our CPU
cluster on another task.

Greg