The format of p3at-YYYYMMDD-HHMMss.log is as follows:
<TIMESTAMP> odom <Current Speed mm/s> <X-position> <Y-position> <Theta> 
<TIMESTAMP> <interface name> <SSID of router> <BSSID of router> <Signal> <Noise>

Before running these two experiments, I calibrated the driving distance.  
Prior to calibration, the robot was driving 20% shorter than perceived.

|- 20080211-T2/ : Contains the logs from driving in a straight line at 100mm/s 
       stopping for 10s at every 4m.  In this case, I used a modified Kismet 
       which output signal strength readings approx every .5s.  It seems 
       that this may affect the accuracy of the signal strength reading.
       Starting point:  (3m, 5.5m) //NOTE: the robot reference pt is (0,0)
       closest router:  router01 (0m, 5.5m)
|- 20080211-T3/ : Drive in straight line from same starting point as T2 but 
       stopped for 20s every 4m.  Here, used the regular kismet version, 
       2007.10.01R1.  The signal strength readings come approx every 1s.
       Starting point:  (3m, 5.5m) 
       closest router:  router01 (0m, 5.5m)