Wow, that is exactly what I am doing: time stamping with 32-bits.
I tried both at CC2420TransmitP and above RadioTimeStamping interface.
Thanks for the valuable information.
Now I have to think about how to handle this...
Thanks a lot.
- jpaek
Philip Levis wrote:
On Nov 4, 2007, at 5:41 PM, Jeongyeup Paek wrote:
To close the thread for now... my hypothesis was wrong.
I went back to
t1/CntToRfm/TOSBase, t2/BlinkToRadio/BaseStation,
on the same testbed and the range seems to be identical.
The main difference between my code and above (I think)
is capturing timestamp at sfd interrupt.
I need further investigation and experiments with my code.
Thanks to everyone for help.
Good to know. Thanks for following up on this. Yeah, interrupt handler
durations are tricky on radio stacks; when Sam Madden was writing
TinyDB, he observed something very similar, and it took him and Wei Hong
about 2 days to figure out the cause was changing timestamps from 16 to
32 bits; the few extra cycles made packet reception drop by something
like 80%.
Phil
--
Jeongyeup Paek
Ph.D. student
Embedded Networks Laboratory
Department of Computer Science
University of Southern California
http://enl.usc.edu/~jpaek
_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
