Abstract: Aiming at the limitations of existing covariance matrix-based blind spectrum sensing algorithms in insufficient utilization of multi-antenna signal correlation and degraded detection performance under low signal-to-noise ratio (SNR) conditions,this paper proposes a fully blind spectrum sensing algorithm based on covariance diagonal-to-nondiagonal ratio detection (CDNRD).The algorithm constructs a detection statistic as the ratio of the sum of the non-diagonal elements of the sample covariance matrix to the total energy of the received signals.By jointly utilizing the amplitude and phase information of the signals,it fully exploits the spatial correlation characteristics of multi-antenna signals.Simulation results demonstrate that in scenarios with low SNR (-15 dB) and strong correlation (correlation coefficient ρ=0.9),the detection probability of the CDNRD algorithm significantly outperforms existing methods,including covariance absolute value detection (CAVD),maximum-minimum eigenvalue detection (MMED),gini index detection (GID),and energy detection (ED).Moreover,the algorithm exhibits robustness and stability across varying sample sizes,antenna numbers,and correlation coefficients.

Key words: cognitive radio, spectrum sensing, blind detection, sample covariance matrix, decision threshold