On Fri, 2007-02-09 at 09:37, Philip Levis wrote: > > Those are all Berkeley TelosB nodes; they are the Omega testbed. > > Section 6.1 reports the variations across the mirage testbed of micaz > nodes. > > We have data for telosb's from a variety of manufacturers (moteiv, > crossbow, berkeley, etc.); I'll ask the student involved to take a > look at the moteiv numbers.
The paper has data for the mirage (micaz) and omega (berkeley telosb) testbeds. Kannan ran the same experiment on the motelab testbed (Moteiv tmote). Here's a link to the plot: http://sing.stanford.edu/data/motelab-chl11-noise-floor-distr.jpg This is on 802.15.4 channel 11. We calculate the noise floor of a node to be the mode of its RSSI readings, in order to filter out things like 802.11 packets and other interference sources. Note that the distribution is going to much broader than the omega one due to the number of nodes involved; omega was 25 or so nodes, while this is a 140 node sample. So you can get more extreme outliers (-91 dBm and -102 dBm). The most common noise floor is -96 dBm, with -100 to -95 dBm all having 10% or more of the samples. Phil _______________________________________________ Tinyos-help mailing list [email protected] https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
