Re: wireless simulation

[John Heidemann <johnh@ISI.EDU>]

On Thu, 16 Sep 1999 10:27:18 +0800, Tan Yuan Sin wrote: 
>Have anyone tested the coding of protocols in NS?
>I could not get the same results when I tested the DSR protocol using the
>same scenario and connection model as in the CMU's comparison of the
>different protocols. How do you test/know whether the simulator is working?

Please see <http://www-mash.cs.berkeley.edu/ns/ns-tests.html>
for a description of what we do and don't test, and how.

There are test cases in the current code (I don't know about the 2.1b5
release) for some of the ad hoc routing protocols, but they don't have
the same level of maturity as (say) the TCP code.

To quote from the above web page:

	You are advised to look at what aspects of the protocols are tested in
	the test suite before drawing research conclusions from these
	protocols.


>To test the protocols (DSR anf DSDV), I modified the wireless.tcl in tcl\ex
>by changing opt(x), opt(y) ,the stop time, the protocol used (dsr or dsdv),
>the scenario file and the connection file accordingly. Is that the correct
>way to run a simulation on wireless network?

You seem to be asking two questions here:
1) how should I test these protocols
2) how should I run a wireless simulation.

What you say sounds like (2).
For (1), you should either compare how it behaves to a real protocol
implementation and/or to the specification.

   -John Heidemann