Hypothesis testing with the identity of high-dimensional covariance matrices for Cognitive Radios

Cognitive Radio technologies are the biggest solution to the increasing spectrum scarcity problem. The most significant problem in Cognitive Radio is to find spectrum gaps both quickly and successfully. A new spectrum sensing model is proposed that applies a unit test to the covariance matrix of the received signal by the Primary Users to overcome this problem. Thus, spectrum decisions can be made quickly with the proposed detector and these gaps can be evaluated. Our proposed detector is quite suitable for both low detection speed and low processing costs. It also offers successful results with different antenna combinations. Besides, the test statistic and threshold for the proposed method were obtained theoretically. However, the performance of the this method was tested in a MIMO-OFDM based detection model. According to the obtained simulation results, the proposed method performed better detection performance than the conventional Energy Detection (ED), Maximum–Minimum Eigenvalue (MME) ,Covariance Frobenius Norm CFN-based detector and other detection methods. Finally, it has been observed that the proposed detector has advantages in terms of processing cost and detection time. • A novel spectrum sensing algorithm based on covariance matrix. • The detection probability for the proposed method was derivated theoretically. • The proposed method is successful according to the studies conducted in recent years. • Performance validation through extensive simulations. • Additionally, the proposed detector can be applied to many hybrid models in further studies.