Meeting notes 05/16/08
* LPL-enabling net2 protocols * TEP4 and documentation
* Razvan, JHU * Andreas, JHU * Phil, Stanford * Om, USC
* CTP+LPL experiment * LPL API discussion * TEP 4 review
Om: We need to "LPL-enable" all net2 protocols, or at least make sure that they do not break when LPL is used.
The current LPL interfaces are per-packet. This is a pain for network protocols. One potential idea is to put a shim on top of the current AMSender that set LPL and otherwise acts as a pass-through.
Experience with running net2 protocols with LPL.
Om: CTP does not seem to work very work over LPL. Need to address that.
Phil: Have you run any LPL tests on tutornet?
Om: we are getting 90%-95% reliability. However we are using a different root (sending every 30 sec). The current LPL setting is broken (1 sec receive interval). (ed. I think Om refered to LPL when he said 'their setting is broken')
Phil: Should run more tests at low-power.
Om: The reason why we get 90-95% with LPL instead of 100% is because queues overflow.
Om: It would be nice to have a coding standard LowPowerSenderAM.nc everything is pass-through but sets LPL. Phil will do it.
Phil: One implementation would need to pass a parameter (length of preamble or RSSI sampling frequency). Problem is that multiple instances might have different intervals. A better solution would to have a central LowPowerControlC that has uniform value and also has receive parameters.
Om will send an email to devel. (ed. done and many emails have followed...)
Om: Dissemination works fine but don't know about DIP
Razvan: Test Deluge with LPL by next week.
Phil: 2.1 release is starting to approach. code freeze at the end of this month. Do fixes sooner rather than later.
Default settings: When one compiles MultihopLQI what would the duty-cycle be? It would be nice to have LPL-enabled versions of the test applications. For example: TestNetwork and TestNetwork.LPL.
Om: TEP4. Change the AM types. Razvan will also take a look at it.
Documentation for protocols.
Razvan quietly added documentation for Deluge.