[olsr-dev] LQ code enhancements

Sven-Ola Tuecke (spam-protected)
Mon Feb 28 18:29:56 CET 2005


Hi Thomas,

Single/Lonley TC? Every UDP packet comes with (can't remember exacly) ~30 
Bytes overhead in the ethnernet frame. That's 1400/10*30 == 4200 Byte 
instead of a single 1400er UDP. Plus! RTS negnotiation for each. Sounds like 
a bad idea. Ah - maybe a max. OLSR-Packetsize may help one day - if the 
packets grow above MTU...

But: What's wrong with considering the lower protocol level when dealing 
with lower protocol level characteristics? LQ is examinging a single UDP 
(hence the ethernet) frame - not the single OLSR message which is one layer 
above.

You will *need* to examine the lower proto levels to get more accurate 
results. Maybe you climb down the protocol ladder and check the cellar (IP 
fragmentation, WLAN Fragmentation, ATM QoS info and the like) for LQ 
enhancements (in 2009 of course ;-)

Just my 2 cents,
Sven-Ola

"Thomas Lopatic" <(spam-protected)> schrieb im Newsbeitrag 
news:(spam-protected)
[---snip---]
> If this is a problem, then we should think about configuration options 
> along the lines of "LonelyHello", "LonelyTc", etc. that force HELLOs, TCs, 
> etc. to be sent in their own individual UDP packet separate from any other 
> messages. This shouldn't be much programming effort. Simply flush the 
> output buffer before adding a HELLO, a TC, etc., add the HELLO, the TC, 
> etc. to the buffer, flush the output buffer again.

[---/snip---] 





More information about the Olsr-dev mailing list