Net2WG/Notes/20090814

From TinyOS Wiki
Jump to: navigation, search

Meeting Notes


Agenda


* bls, blip release schedule/status
* dissemination protocols
* ctp bugs/cleanup

Participants

* Om, USC
* Razvan, JHU
* Thanh,PDX
* Martin, Crossbow
* Kaisen,UCSD

Action Items


Discussion Notes


. [General discussion about 2.1.1 release...]

[BLS] - schedule to release with 2.1.1 in the end of August

[DHV]

Thanh : DHV is a dissemination protocol supporting network reprogramming. It's purpose is similar to DRIP and DIP. We evaluated with TOSSIM and 1-hop network, DHV is about 50% better than DIP and has similar memory usages compared to DIP. It hasn't been compared with DIP on real multi-hop testbed though. We are trying to evaluate DHV using Motelab.

Martin  : It should be tested with multihop before checking in

Thanh  : Ok, I will do that

[DIP]

Om  : Thanh has some problems trying to get DIP to work with his testbed

Kaisen  : Ok, what node are you using?

Thanh  : MicaZ, I'm not sure if it's the problem with DIP or our testbed. We can simulate it though.

Kaisen  : Does it not run...?

Thanh  : It doesn't send the expected messages. It transmits summary messages, but not the vector messages... ...

Kaisen  : I guess if you can send me the code you are testing, I can try to make it work

Thanh  : Ok, I can also send you the log, we use one basestation to listen to all messages, I can send you that log too

Kaisen  : Ok, that would be helpful

....

Om  : Thanh, You might want to test DHV with LPL too

Thanh  : Oh ok, most of the time LPL is disabled in DHV. I will check that.

...

Om  : Razvan, is LPL working with other dissemination protocols?

Razvan  : DRIP is working for sure, I don't remember if DIP is working

Om  : That is something interesting to check...

Kaisen  : Ok

[CTP]  : tested and checked in ...

Om  : There is problem with link estimator overhead and forwarding engine overhead...

Martin  : Can you send me the CTP log?

Om  : Yes, I will do that

Om  : There is also a problem sending ETX for other groups, checking gradient of zero...

Martin  : That is not a bug... I will take a look at the log you send me.

[Some discussion about availability the latest code tree, we are a little behind the schedule, predicting Sep for a final for the community]

That's all I got, please let me know if I mixed up the names because it was the first time I called in.