Thanks for the reply, but why the range of the congestion delay is shorter than that of initial delay? Isn't it natural to increase the range of backoff delays when the channel is not clear?
Philip Levis wrote: > > On Jun 12, 2008, at 12:29 PM, Anton wrote: > >> Hello! >> >> There is probable bug in the >> /tinyos-2.x/tos/chips/cc2420/csma/CC2420CsmaP.nc (lines 218 and 225) >> >> Communicating wireless modules should wait some random delay before >> performing Clear Channel Assess. >> According to the IEEE 802.15.4 standard, that delay should be >> random(2^BE-1) backoff periods, where BE can vary from 3 to 5 (the >> default values macMinBE and macMaxBE, IEEE 802.15.4-2006, pages >> 163-164). This means that for CC2420 the maximum backoff delay is 31*320 >> = 9920 microseconds (when BE=5), and the maximum initial backoff delay >> is 7*320 = 2240 microseconds (since initially BE=3). >> >> But according to CC2420CsmaP.nc file the maximum initial backoff delay >> can be around 9765 microseconds (line 218) and maximum congestion >> backoff delay can be around 2441 microseconds (line 225). >> It seems that the ranges of these random delays are swapped. Even in >> TinyOS 1.x the range of the initial delay is smaller than that of >> congestion backoff delay >> (/tinyos-1.x/tos/lib/CC2420Radio/CC2420RadioM.nc, lines 745 and 752). In >> TinyOS 2.x this is conversely. >> >> Also, according to the standard the range of the random delay should >> expand (by default BE should change from 3 to 5) each time the channel >> is sensed busy (CSMA/CA algorithm - page 172 of the standard). But the >> range of all random delays is constant both in TinyOS 1.x and TinyOS >> 2.x. >> >> Please, tell whether it is bug or not. > > No, it is not a bug. TinyOS does not have an 802.15.4 MAC. > > Phil _______________________________________________ Tinyos-help mailing list [email protected] https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
