Stopping Criteria for Iterative Compressive Sensing of Sparse Angular-Frequency Spectrum

Recent trends have shown that spectrum regulators are opening up more shared spectrum bands for fifth generation mobile communications system (5G) and Beyond 5G (B5G) networks. The possibility for spectrum sharing in the angular domain as well as the frequency domain simultaneously increases with the use of array antennas in 5G systems. An efficient technique for joint sensing of the angular-frequency spectrum called the Wideband Group Simultaneous Orthogonal Matching Pursuit (WGSOMP) algorithm was previously proposed based on compressive sensing (CS) assuming a sparse spectrum. In this paper, several stopping criteria is proposed for the WGSOMP algorithm, which aims to maximize the probability of detection while minimizing the probability of false alarm in unoccupied bands. Performance of the proposed stopping criteria is evaluated through Monte Carlo simulation, and results show that the probability of false alarm can be reduced by more than half without reducing the probability of detection at SNRs larger than 0 dB, especially when the transmit signal parameters are not confined to the angular-frequency grid defined in the overcomplete dictionaries.