frequency offset

[dac4755]

Hi. I worked through the MAC spoofing example and am finding that the frequency offset numbers I see are not what I expected. How consistent will those numbers be? For very small samples I see numbers that might be used to identify a device but if I collect for a longer period of time with more devices present, I don't see frequenct offsets precise enough to uniquely distinguish between all devices. To uniquely identify a transmitting device do I need to take an average reading over time? For example, for one device I saw:

64:5d:86:65:9b:04 -10744.2077295474
64:5d:86:65:9b:04 113.754962641629
64:5d:86:65:9b:04 -11574.7479122258
64:5d:86:65:9b:04 -12045.783151968
64:5d:86:65:9b:04 -12093.2101112646
64:5d:86:65:9b:04 -12618.2401419467
64:5d:86:65:9b:04 -13685.6061654755
64:5d:86:65:9b:04 -214.024637241781
64:5d:86:65:9b:04 -2151.75582665642
64:5d:86:65:9b:04 -3864.59603403644
64:5d:86:65:9b:04 -4404.2917763495
64:5d:86:65:9b:04 -5038.52738640282
64:5d:86:65:9b:04 -606.970067033554
64:5d:86:65:9b:04 -6339.94819204653
64:5d:86:65:9b:04 -9330.75877775144

At the end of the blog post the Future section says you can get a better estimation from the frame_equalizer block. What is the correct way to set that up? I tried using blocks in a similar fashion to getting measurements off the WIFI Decode MAC block but that didn't seem to work at all. I'm using an N210, fwiw.

thanks