Efficient Joint Sensing of Sparse Angular-Frequency Spectrum based on Compressive Sensing

Abstract

It is expected that the highly-dense deployment of private fifth generation mobile communications system (5G) microcells will be further accelerated in order to meet the demands of a wide range of use cases in various environments. This motivates the need for spectrum sharing research in the mmWave spectrum. The use of array antennas in 5G systems opens the possibility for spectrum sharing in the angular domain in addition to the frequency domain. This paper investigates the use of compressive sensing (CS) to efficiently perform spectrum sensing over a wide bandwidth in the frequency and angular domains simultaneously using only a small number of compressed measurements. The CS problem is formulated using the Kronecker product of overcomplete dictionaries, where the overcomplete dictionary in the angular domain consists of frequency-dependent steering matrices. Modification of the Kronecker Orthogonal Matching Pursuit (Kron-OMP) algorithm is proposed, and group-sparsity is enforced over the frequency domain to improve detection of wideband signals. The effectiveness of the proposed approach for wideband spectrum sensing is confirmed through Monte Carlo simulation.