Compressive RF sensing using a physical source of entropy


Compressive Sensing is a promising new field that has unlocked novel devices like the single pixel camera. In the world of RF communications, however, experiments with compressive sensing have yet to catch up with the theory; most demonstrations of RF compressive sensing involve a high speed clock somewhere in the signal chain, diminishing the advantages of slow and brief sampling that compressive sensing offers. We present our initial results from the first compressive RF sensor architecture that does not require a high-speed clock anywhere in the signal chain. By employing a physical source of entropy and applying a sub-clock resolution timing scheme to record randomly-timed digital samples, we demonstrate a compressive RF sensor concept that, when extrapolated to higher bandwidths, can significantly reduce signal measurement energy requirements.

2011 IEEE Statistical Signal Processing Workshop (SSP)