Thank you for all the advice. I am using the modified Collection protocol from the tutorial. I'm still not sure of how it gathers a tree from the program, but I used the Listen function and found that each packet received at the root will include the id of a mote it has found since being turned on. Is there a better way to trace the hops? Only the BaseStation mote is connected via USB to the PC, the others use batteries. Any suggestions for the power adjustment on the radios?
Thanks, Ellen On Wed, Jul 29, 2009 at 1:13 PM, Philip Levis <p...@cs.stanford.edu> wrote: > > On Jul 29, 2009, at 7:10 AM, Ellen Shlossberg wrote: > > Hello All, >> >> I have been reading through the mailing list and relative papers on >> multi-hop routing in TinyOS and I am hoping someone can suggest a good and >> simple implementation. I am using tinyos-2.x with mica2 motes and the >> options for routing aren't always applicable to both 2.x and mica2. This is >> a document I found useful on describing the routing types, but it does not >> specify the locations of files and how to implement the protocols. >> >> http://db.lcs.mit.edu/madden/html/tinyos-nsdi04.pdf >> >> My goal is to have around 20 motes turn on, discover routes to each other >> by talking to their one-hop neighbors, and then maintain this network and >> notice if any new mote has entered or if an existing one has left. It >> appears that AODV would be useful, but the routes would have to be retained >> in a routing table for an extended period (not just for one-time use). What >> would be the best routing method to use? >> >> I have Dissemination and Collection working, but I believe they do not >> maintain specific routes, but rather used route discovery each time they >> need to disseminate or collected. I've also read about multiHopRoute - but >> could not find a specific implementation. Same for LQI and CTP. Do these >> work with mica2? Is there a simple way to test these out in a small network >> of motes - say 4 or 5 motes? >> > > CTP works on mica2 nodes. Here's a paper describing how it works: > > http://sing.stanford.edu/pubs/sing-09-01.pdf > > CTP does not use on-demand route discovery: it continuously maintains > routes to collection points. So it's more like DSDV than AODV/DYMO. > > Phil > -- Ellen Shlossberg Columbia University MS Electrical Engineering
_______________________________________________ Tinyos-help mailing list Tinyos-help@millennium.berkeley.edu https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
