Double threshold based spectrum sensing technique using sample covariance matrix for cognitive radio networks

Abstract

The spectrum sensing is a vital stage in cognitive radio networks. The performance of energy-based spectrum sensing scheme degrades under noise uncertainty, which can be improved by covariance-based sensing methods. In this paper, a double threshold-based spectrum sensing scheme using sample covariance matrix of received signal is presented, that improves the performance of conventional covariance-based spectrum sensing methods. The two thresholds are derived using random matrix theory. The performance of proposed scheme is verified through simulations.