Compressive techniques for wideband frequency hopping signals

Abstract

Compressive Sensing (CS) is an emerging theory that has a lower rate of signal acquisition as compared to the renowned Nyquist-Shannon sampling theorem. “CS theory asserts that one can recover from far fewer samples or measurements than traditional method uses” [1] With a lower amount of measurements taken as compared to the Nyquist rate, the complete signal is then subsequently reconstructed. There are an abundant of sub-Nyquist sampling schemes available for various purposes but multi-coset sampling (MCS) was selected in this project as it shares many common characteristics to other sub-Nyquist CS sampling protocols. This report is targeted at providing a summary of the MCS techniques that can be utilized for wideband frequency hopping (FH) communication signals. The project focuses on MCS with FH signals experimenting with different variables to achieve the lowest possible NMSE to improve the recovery performance.