On the design and throughput analysis of a new MME detector using Bartlett´s method

Detection of primary users is a very crucial task for cognitive radio networks. On the other hand, the choice of the spectrum sensing algorithm affects the achievable throughput of the secondary network when the cognitive radios employ the interweave approach. Focusing on the eigenvalue-based detector, it can be seen that most proposed methods are based on time domain. Therefore, in this paper we present and analyse a new class of eigenvalue detectors in frequency domain (FD). The major contributions of this paper are of two parts. First, we introduce a new minimum-maximum eigenvalue (MME) spectrum sensing method based on Bartlett´s method of estimation. By making use of the eigenvalues of the diagonalized quadratic form representation of Bartlett´s estimate, the distribution of the ratio of extreme eigenvalues is studied, and then the performance of the detector is addressed and accurate expressions where obtained for the probabilities of false alarm and detection. Second, assuming that the secondary network employs an interweave approach we study the average achievable throughput. All obtained formulas are verified through Monte Carlo simulations. The obtained results gives an insight into the validity of using MME detection in FD.