Photonic-Assisted Compressive Sensing with Dispersion Fiber

The technique of microwave photonic filtering and the structure of dual-drive Mach-Zehnder modulator (DDMZM) are applied to photonic-assisted compressed sensing (CS) for better system performance. The scheme involves modulating the input radio frequency signal and pseudo-random bit sequence (PRBS) onto a single-wavelength optical carrier and propagating them through the dispersion fiber. Low-pass filtering required for CS is achieved based on frequency-dependent power degradation caused by double-sideband modulation in a dispersive optical link. To maintain the zero-mean property of the observation matrix in the photonic CS system, the DDMZM is employed to perform frequency mixing between the input signal and a bipolar random sequence. Digital domain nonlinearity compensation is implemented to avert the errors introduced by the small signal approximation. Compressive sensing has extensive application demands in communication, radar, and other fields. Simulation and experiments verify the feasibility and potential of the proposed scheme.