Yesterday I gave a talk on the Village Telco to a graduate computer
science class at the University of the Western Cape here in South
Africa. Got some great feedback and interesting questions from the
students. Their professor, Bill Tucker, is keen to involve them in
the Village Telco project. One area of interest he has is in VoIP
over Mesh WiFi performance. He wondered whether we had done any
simulations using NCTUns (http://nsl10.csie.nctu.edu.tw/). Has anyone
used this simulation software? Would it be useful for testing and
optimising the MP firmware?
Cheers... Steve
--
Steve Song
Telecommunications Fellow, Shuttleworth Foundation
email: steve...@shuttleworthfoundation.org
work: +27 21 970 1220
mobile: +27 83 482 2088
skype: steve_l_song
blog: http://manypossibilities.net
twitter: stevesong
On Tue, Mar 2, 2010 at 11:21, Steve Song
<steve...@shuttleworthfoundation.org> wrote:
> Yesterday I gave a talk on the Village Telco to a graduate computer
> science class at the University of the Western Cape here in South
> Africa. Got some great feedback and interesting questions from the
> students. Their professor, Bill Tucker, is keen to involve them in
> the Village Telco project. One area of interest he has is in VoIP
> over Mesh WiFi performance. He wondered whether we had done any
> simulations using NCTUns (http://nsl10.csie.nctu.edu.tw/). Has anyone
> used this simulation software? Would it be useful for testing and
> optimising the MP firmware?
I haven't used this exact simulator, but in general I'd say it's a very
handy to be able simulate different network conditions and (may be
even more important) to reproduce once-recorded tracks with interesting
artifacts. This is an essential when you work with e.g. jitter buffer and
PLC algorithms, as you can't rely that your local network will have all
that delays people see in real deployments ;) So at SIPez we developed
our own simulator which ran on NSLU2's (with USB-Ethernet as second
Ethernet port) and it worked great.
--
Regards,
Alexander Chemeris.