Compressive sensing with a microwave photonic filter

In this letter, we present a novel approach to realizing photonics-assisted compressive sensing (CS) with the technique of microwave photonic fi ltering. In the proposed system, an input spectrally sparse signal to be captured and a random sequence are modulated on an optical carrier via two Mach – Zehnder modulators (MZMs). Therefore, the mixing process (the signal to be captured mixing with the random sequence) is realized in the optical domain. The mixed optical signal then propagates through a length of dispersive fi ber. As the double-sideband modulation in a dispersive optical link leads to a frequency- dependent power fading, low-pass fi ltering required in the CS is then realized. A proof-of-concept ex- periment for compressive sampling and recovery of a signal containing three tones at 310 MHz, 1 GHz and 2 GHz with a compression factor up to 10 is successfully demonstrated. More simulation results are also presented to recover signals within wider bandwidth and with more frequency components.