Re: [B.A.T.M.A.N.] Masters thesis

[Marek Lindner <lindner_marek@yahoo.de>]

Hi,

> Was the protocol implemented yet on a simulation platform like omnet?
> For comparing purposes (for example with OLSR) it could be interesting
> to simulate the protocol.

there is no simulator support that I know of.


> Further the package load will be tunnelled through batman network by an
> udp tunnel. There was reasons to do this. But why we can not establish
> an tcp-tunel on top of the existing tunnels. So earn more overhead but
> the delivery of the related packages are granted.

It is not necessary to implement the tunnel with TCP as all protocols that use 
the tunnel already have a mechanism to detect packet lost. If you issue a HTTP 
query over a tunnel it will look like this:
HTTP -> TCP -> UDP Tunnel -> Internet

Your setup would look like this:
HTTP -> TCP -> TCP Tunnel -> Internet

Next to the overhead which comes with TCP you will face a double congestion 
window (see TCP congestion avoidance). In a lossy environment TCP over TCP 
connection will run at the lowest speed possible without a chance to recover. 


> In respect to IPv6: What is the problem with IPv6? The batman protocol
> self should work with ipv6? As payload it could be realized by an tunnel
> too.

Its not a problem - just somebody needs to do it.  :-)

Regards,
Marek