[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [ns] How long should a simulaion be ?

To: Andrea Josi <[email protected]>
Subject: Re: [ns] How long should a simulaion be ?
From: John Heidemann <[email protected]>
Date: Tue, 21 Aug 2001 15:25:27 -0700
cc: [email protected]
In-Reply-To: Message from Andrea Josi <[email protected]> of "Thu, 16 Aug 2001 10:00:50 +0200." <[email protected]>
Sender: [email protected]

>Hi,
>
>i have to simulate the transmission of CBR traffic
>among several clients and servers and calculate the
>end to end delay variance.
>I will use Pareto traffic sources as background
>traffic:
>How much should my simulations last to bring
>meaningful results ?
>
>Thanks !

Long enough to avoid transient effects.

(There is no simple answer.
You need to look at the output and determine when you've reached
steady state.)

   -John Heidemann

References:
- [ns] How long should a simulaion be ?
  - From: Andrea Josi <[email protected]>

Prev by Date: [ns] Question on GT-ITM topology generator
Next by Date: Re: [ns] unable to find old version of ns -- pls help
Prev by thread: [ns] How long should a simulaion be ?
Next by thread: [ns] Problem of Test-suite-IntServ.tcl
Index(es):
- Date
- Thread