A modified eigenvalue-based cooperative spectrum sensing algorithm

Abstract

Spectrum sensing is an essential function in cognitive radio (CR) to dynamically detect temporarily unused frequency bands, which improve the efficiency of utilization of frequency spectrum. In this paper, based on traditional maximum and minimum eigenvalue detection (MME), a modified algorithm is presented by proposing a new decision statistic and the corresponding decision threshold is also deduced based on Random Matrix Theory. The modified algorithm is robust against noise uncertainty and has good performance in low SNR for the nature of correlation of primary signal. The modified algorithm is simulated in MATLAB and results are analyzed. Compared with traditional MME algorithm, the modified algorithm has better performance in detecting probability.