Sparse Spectrum Recovery of Streaming Signals Based on Multi-Resolution

In this paper, we consider the problem of multi-resolution analysis for sparse spectrum of streaming signals. The multi-resolution compressed sensing for streaming signals (MRCSSS) is proposed, which fully exploits the inner relationship between frequency resolution and sensing time based on the analog-to-information converter(AIC). Different from the existing algorithms, we deduce the correlation between frequency support of high-resolution and low-resolution spectrum. Then, the recovered frequency support of low-resolution spectrum is utilized to estimate the frequency support of high-resolution spectrum, which serves as a priori knowledge for a more efficient reconstruction of high-resolution spectrum. Simulation results have demonstrated the effectiveness of our algorithms.