[sim-wg] TOSSIM Power Modeling

[Tal Rusak]

To the Simulation Working Group:
     Over the past few months, I have been working on collecting long-term
signal RSSI traces, to be used for modeling power in TOSSIM. To
facilitate this process, I have written a program that collects RSSI
data from a network of two motes, sending and receiving packets. In
addition, I have collected several traces (see
RadioCountToLedsSerial/report.pdf in the attached ZIP).
     Specifically, the program collects data about both reception RSSI as
well as noise (see lines 138-152 and 163-172 of
RadioCountToLedsSerial/RadioCountToLedsC.nc). Noise is collected
using a technique similar to the RSSISample application developed at
Stanford; I am not sure what the time delay is between the collection
of packet RSSI and noise using this strategy. One point that concerns
me is that in some of these traces (see
RadioCountToLedsSerial/Traces), RSSI = noise. However, since RSSI is
the linear sum of noise and power, this means that the power was
lower than the recorded noise (which should not have been received in
the first place). I am not sure if this is a result of the time delay
between RSSI and noise samples or another reason.
     I have also implemented a variant of TOSSIM that models power based
on the collected traces, please see
RadioCountToLedsSerial/CPMPowerReport.pdf for more details.
     Please let me know if you have any comments or suggestions about the
RSSI collection and/or the modeling algorithm. Have a happy new year!

Thanks,
Tal
-------------- next part --------------
A non-text attachment was scrubbed...
Name: RadioCountToLedsSerial.zip
Type: application/zip
Size: 1787746 bytes
Desc: not available
Url : https://www.millennium.berkeley.edu/pipermail/sim-wg/attachments/20071230/498e6941/RadioCountToLedsSerial-0001.zip