Changes between Version 1 and Version 2 of WARPLab/Examples/Spectrogram

Timestamp:

Jan 15, 2015, 1:01:09 PM (9 years ago)

Author:

chunter

Comment:

--

WARPLab/Examples/Spectrogram

@@ -9 +9 @@
 As of WARPLab 7.5, the available size of the receive sample buffer is 128M samples for the 2RF design. This is 4000x larger than the 32k sample buffer sizes in WARPLab 7.4. For more perspective on that number, 32k samples at 40 Msps (i.e. 40 MHz of bandwidth) yields an entire buffer length 819 µsec. This is long enough for a reasonable length packet transmission, but not much more. With WARPLab 7.5, capturing the entire receive buffer yields a full 3.35 seconds of reception at the same 40MHz of bandwidth.
 
-There are many usage scenarios where these expanded buffer lengths are useful. One such usage scenario is monitoring channel activity. With over 3 seconds of received waveform at a full 40MHz of bandwidth, we can see both a large extent of channel activity at very fine timescales. One popular visualization of frequency content across time is known as a [http://en.wikipedia.org/wiki/Spectrogram spectrogram]. In this example, we will use this technique to visualize activity in the 2.4 GHz band.
+There are many usage scenarios where these expanded buffer lengths are useful. One such usage scenario is monitoring channel activity. With over 3 seconds of received waveform at a full 40 MHz of bandwidth, we can see a large extent of channel activity at very fine timescales (25 ns samples). One popular visualization of frequency content across time is known as a [http://en.wikipedia.org/wiki/Spectrogram spectrogram]. In this example, we will use this technique to visualize activity in the 2.4 GHz band.
+
+MATLAB offers a sophisticated [http://www.mathworks.com/help/signal/ref/spectrogram.html spectrogram tool] as part of the Signal Processing Toolbox. Our simple implementation of a spectrogram in this example does not offer the same advanced features like windowing or FFT overlap. Instead, our implementation works by simply plotting the output of {{{M}}} sequential FFTs of {{{N}}} samples each.