An Effective Spectrum Sensing Method via Space-Time Averaging Technology

Spectrum sensing is a key technique in the cognitive radio. It enables secondary users to detect spectrum holes and monitor primary users´ activity. However, many factors make spectrum sensing much difficult, such as very low signal-to-noise ratio, and noise uncertainty, and quite short sensing time, etc. In this paper, a uniform linear array is configured for each secondary user´s receiver. The sample covariance matrix can be effectively modified by the Hermitian and Toeplitz property. Thus, the primary users´ signals can be much more accurately detected by comparing the ratio of the maximum eigenvalue to the average of others of the modified sample covariance matrix of received signals with a threshold. Simulation results not only show the proposed method can efficiently overcome the influence of the aforementioned negative factors, but also indicate that it need no knowledge of primary users´ signals and noise in advance.