Blind and Sparsity Level Adaptive Reconstruction of Wideband Time-frequency-varying Signals with Sub-Nyquist Sampling

Abstract

This paper addresses the problem of wideband time-frequency-varying signal sub-Nyquist sampling and reconstruction based on compressed sensing (CS) framework. We propose a system of blind and sparsity level adaptive signal reconstruction for wideband signals with sub-Nyquist sampling. We utilize modulated wideband converter (MWC) that deals well with multi-band signals to acquire sub-Nyquist samples, change the signal sensing and reconstruction model to parameters estimation model in the array signal processing, and apply iterative adaptive approach (IAA) to recover spectral support and reconstruct signals simultaneously without any prior knowledge. Simulation results show that the proposed method outperforms the continue to finite (CTF) following MWC in low signal-to-noise ratio (SNR).