[Numbat] Types of traffic supported

[Andrew Lapidas]

Actually, I was unable to find anything else decent concerning WiMAX
and omnet.  And you are correct, I very quickly noticed the lack of
comments! I said I had given up on this project - the professor has
let me go another route not involving a simulation at all, but an
Android app instead.  Also, I am a dedicated Linux user, but I was
using your simulation under Windows.  You are correct, I am from the
University of Massachusetts Amherst.

-Andy

On Mon, May 5, 2008 at 3:26 AM, Tomasz Mrugalski <thomson at klub.com.pl> wrote:
> On Sun, 4 May 2008 somebody known as Andrew Lapidas wrote:
>
>  > from the base station.  After many hours of reviewing your code, it
>  > seems it's a little over my head (this is only a half semester long
>  > project).  Also, it seems your simulation is concerned more with IPv6,
>  I'm focusing on handover efficiency, actually. The whole handover process:
>  WiMAX handover, dynamic IPv6 configuration (especially DHCPv6, as I'm also
>  author of Dibbler - a DHCPv6 implementation) and IP mobility (Mobile
>  IPv6).
>
>  From my point of view, WiMAX congestion and level of service is only small
>  part of the whole system. Anyway, here are some guidelines:
>
>  - there is a fixed number of SSes. Each SS has its corresponding node in
>  the Internet. (number of SSes is specified omnet.ini: physim.numSS = 3)
>
>  - each SS may be fixed(0) (i.e. not moving), performing handover after
>  specific timeout(1) or truly mobile (2) (i.e. moving and assessing signal
>  strength to neighboring BSes). Type of mobility is specified by
>  physim.SS[0].wmaxIsMobile = 1 in omnet.ini
>
>  - Traffic is generated by both SS and its corresponding node. See
>  CN[0].IPv6Node or SS[0].ssIPv6.IPv6Node for module that generates actual
>  traffic. It's very simple. After initial delay (InitialDelay parameter) it
>  starts to send bursts of packets (number of packets is randomized from 1
>  to BurstSize). Each packet may be of size 1 to PacketSize. Next burst is
>  sent after BurstInterval. The (very simple) code is locaed in
>  ipv6/ipv6node.cc (see void IPv6Node::generateTraffic() method).
>
>  - Please note that IP layer generator/receiver is completely separated
>  from WiMAX traffic class (as it is in a real life). As I don't have any
>  particular need for specifying WiMAX connection parameters, it is
>  hardcoded currently. See wimax/wmaxctrl.cc, method
>  WMaxCtrlSS::createNewFlowEvent(). Currently 2 values are supported:
>  WMAX_CONN_TYPE_BE and WMAX_CONN_TYPE_UGS.
>
>
>
>  > which I am not really concerned with.  Just out of curiosity though,
>  > could you point me to where the traffic generation happens?  I was
>  > unable to find it and I was curious what you are using to generate it.
>  I'm developing Numbat for my Ph.D. As it takes longer than expected, I'm
>  trying to skip unnecessary tasks (read: don't write docs, don't document
>  the code too well).
>
>
>  > Either way, thanks for your reply, it is much appreciated.  Good luck
>  > with your further development!
>  You're welcome. You've said that you gave up. Does it mean you found
>  something better/easier? Is there much choice regarding WiMAX simulations?
>  I've checked about 1.5 year ago and the set of freely available sims was
>  minimal. So I decided to start my own project.
>
>  Unfortunately to you, all Numbat participant were from Poland: me and 2
>  M.Sc. students. So almost everything we've talked about on the number list
>  was in polish. But I suppose it wouldn't very useful anyway.
>
>  I've just uploaded today's snapshot.
>  http://klub.com.pl/projects/numbat/snapshots/ To compile it, type:
>  ./rebuild-makefiles
>  make
>
>  WiMAX sim is in the wimax directory. I'm assuming that you are using Omnet
>  3.3 or later under Linux. If you are Windows user, there's a doc called
>  Omnetpp_win_inst_guide.pdf Be warned, it may be out of date and I won't be
>  able to help you as I don't do Windows anymore :)
>
>  If you have any other questions, feel free to ask.
>
>  Tomek
>
>  p.s.
>  Where are you from? University of Amherst, I suppose?
>
>
> _______________________________________________
>  Numbat mailing list
>  Numbat at klub.com.pl
>  http://klub.com.pl/cgi-bin/mailman/listinfo/numbat
>



-- 
Andrew Lapidas
email / alapidas at student.umass.edu
phone / 774.279.4740