I think you also need to change NETSYNC_DUTYCYCLE in "sync" with NETSYNC_PERIOD_LOG2. I haven't tried it but but did do some time recently trying to understand the moteiv code.
There could also be a problem in NetSyncM.nc::Receive.receive() circa line 243 where a limit on clock skew adjustment of "0x2f" is enforced. This value works out to about a +/-1.5ms window every 2 sec or .15% which seems way too low (and will be halved if you double the PERIOD). MS Timothy Strongton wrote: > Dear Members, > > - Objective: in Moteiv(now Sentilla)'s Delta, try to change the period > between listens from 2 seconds to 4 seconds > - Source: recent Boomerang from TinyOS CVS (in tinyos-1.x/contrib/boomerang) > - Patches applied: > http://sing.stanford.edu/klueska/tos_wiki_files/NetSyncM.nc > - Configuration changes: added "NETSYNC_PERIOD_LOG2=17" to Makefile > - Symptoms: > . On each run, some nodes keep sending data with no loss, but some > nodes only get a little bit of data in sporadically. > . If I change the code(in CC2420SyncSaneP.nc) so that an LED is turned > on while the RF chip is turned on, you can see that the nodes that fail > to send in data are exactly 2 seconds out of sync with the sink > node(a.k.a base station). Sync is perfectly fine when > NETSYNC_PERIOD_LOG2=16 (the default that gives a 2 second period): > actually tested with 25-node configuration that forces multi-hop. > 4-second period doesn't work even when everything is single-hop. > > Thank you very much for your valuable time. > > Timothy > > > ------------------------------------------------------------------------ > > _______________________________________________ > Tinyos-help mailing list > [email protected] > https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
