A Novel Wideband Spectrum Sensing Algorithm for Cognitive Radio Networks Based on DOA Estimation Model

In this paper, we consider wideband spectrum sensing problem in cognitive radio networks. We propose a new method to sense the spectrum based on DOA (direction of arrival) estimation model. In order to sense the spectrum, we examine the MUSIC-like algorithm using fourth-order cumulants matrix of the received signals. Wideband spectrum is assumed to be linear and uniformly spaced among subchannels. To estimate the number of occupied subchannels, we use the singular value decomposition of the fourth-order cumulants matrix and to estimate the location of occupied subchannels, we investigate optimizing a cost function that is obtained from MUSIC-like algorithm. The performance of the proposed algorithm is investigated through simulations. The results indicate the effeciency of the proposed method.