Abstract: Based on the difference between the maximum and minimum eigenvalues (DMME) of the sample covariance matrix,to improve the setting accuracy of sensing decision threshold,a new method is proposed to calculate the threshold by using the latest results of limiting eigenvalue distribution of Wishart matrix.Compared with the traditional DMME algorithm,the decision threshold setting of the proposed algorithm is more accurate.While improving the reliability of decision results,it improves the detection performance of the algorithm with limited number of samples.Both the theoretical analysis and simulation results verify the effectiveness of the proposed DMME algorithm.

Key words: cognitive radio, spectrum sensing, Wishart matrix, limiting eigenvalues, Tracy-Widom distribution