Re: [ns] ns optimizations (wireless)

[John Heidemann <johnh@ISI.EDU>]

Kun-chan writes:
>There are some general tips on ns webpage for
>improving performance of large simulation.
>You might want to take a look if you haven't.

>On Mon, 20 Aug 2001, Dan Iregren wrote:
>> ...
>> The channel layer is quite time consuming, it seems as if about 1/3 of the 
>> simulation time is spent there. I would have guessed a lot more. What other 
>> places would be relevant to examine further?

It would be interesting to do a simple, minimal wireless MAC and
propagation model that perhaps ran faster.  (I don't know how much
this would save.)

Also, Ya Xu explored "gridkeeper" for wireless.  The idea was to
reduce packet replication by bounding the maximal propagation
distance.  Packets are only propagated within "grids" (more or less).
The basic idea is good, but the bookkeeping makes things only pay off
for large scenarios.  Reviving this idea might be interesting.

   -John Heidemann