Re: [ns] goodput calculation

[Lloyd Wood <l.wood@eim.surrey.ac.uk>]

On Tue, 12 Jun 2001, pawan k gupta wrote:

> to try and calculate goodput, isn't it enough to just find out the
> largest packet number that was acked by the tcp receiver and
> divide it by the total number of packets transmitted by the
> sender.

No. (and that really messes up with SACK).

That seems to be aiming for a ratio of effective throughput to packets
sent, or goodput to initial throughput, if you like.

Also, in TCP, packets can be variable-sized (anything less than path
MTU), although ns is a bit skimpy in simulating that.

If you want goodput, you really have to track the left edge of the
window at the receiver as it passes data to the application - and
ndatapack_ isn't at the receiver.

L.

> i found that two such parameters already exists in the ns. these are ndatapack_ and ack_.
> according to this goodput of the connection is
> 
> (ack_ - 1)/ndatapack_
> 
> please correct me if i am wrong.

<L.Wood@surrey.ac.uk>PGP<http://www.ee.surrey.ac.uk/Personal/L.Wood/>