Compressive Sensing Based on Mesoscopic Chaos of Silicon Optomechanical Photonic Crystal

Abstract

Compressive sensing (CS) is an effective technique that can compress and recover sparse signals below the Nyquist-Shannon sampling theorem restriction. In this study, we successfully realize CS based on the mesoscopic chaos of an integrated Si optomechanical photonic crystal micro-cavity, which is fully compatible with the complementary metal-oxide-semiconductor (CMOS) process. Using the sensing matrix, we tested one-dimensional waveforms and two-dimensional images. The ultimate recovery curves were determined by comparing the chaotic sensing matrix with the Gaussian, Toeplitz, and Bernoulli matrices. Our results could pave the way for future large-scale implementations of high-speed CS processes based on fully CMOS-compatible Si-micro-cavities.