Scaled-Energy Based Spectrum Sensing for Multiple Antennas Cognitive Radio

In this paper, for a spectrum sensing purpose, we heuristically established a test statistic (TS) from a sample covariance matrix (SCM) for multiple antennas based cognitive radio.The TS is formulated as a scaled-energy which is calculated as a sum of scaled diagonal entries of a SCM; each of the diagonal entries of a SCM scaled by corresponding row's Euclidean norm.On the top of that, by combining theoretical results together with simulation observations, we have approximated a decision threshold of the TS which does not need prior knowledge of noise power and primary user signal.Furthermore, simulation results -which are obtained in a fading environment and in a spatially correlating channel model -show that the proposed method stands effect of noise power mismatch (non-uniform noise power) and has significant performance improvement compared with state-of-the-art test statistics.