[net2-wg] Packet interface

Philip Levis pal at cs.stanford.edu
Wed Nov 30 16:50:59 PST 2005


On Nov 30, 2005, at 4:25 PM, Kyle Jamieson wrote:

> While most sensornet communication is indeed to a single sink, seems
> like we're restricting apps' flexibility quite a bit with a one
> destination per pool entry policy.  What was the rationale behind not
> using per-packet destinations?

TDMA. You have an opportunity to send N packets to a single  
destination. Also, CSMA+LPL, where a long preamble wakes up a node,  
and others may go back to sleep once they hear the packet isn't for  
them.

We could just say there's a big pool you toss individual packets into  
and the packet pool schedules individual packet transmissions, but  
then this requires a component to commit to (allocate) all of those  
packets beforehand.

It's definitely the case that an app can have N entries, if it wants.  
There's no restriction in that regard. Chances are a pool entry is  
~20 bytes at most.

Phil

-------

"We shall not cease from exploration
And the end of all our exploring
Will be to arrive where we started
And know the place for the first time."

- T. S. Eliot,  'Little Gidding'





More information about the net2-wg mailing list