Técnicas de sensoriamento espectral paralelo à transmissão baseadas em autovalores sob cooperação espaço-temporal em redes de rádios cognitivos

Abstract

Maximizing the secondary cognitive radio network throughput under the restriction of avoiding interference to the primary network is a main concern in recent research. In the classic approach, a specific time interval is designed to the task of spectrum sensing, which penalizes the secondary network throughput. Aiming at a higher throughput than in the classic approach, the continuous sensing mode can be adopted. In this mode, secondary receiving nodes sense the spectrum while other secondary users simultaneously transmit in the same frequency band. In this work it is compared, under the continuous sensing approach, the performances of centralized spatiotemporally cooperative spectrum sensing techniques. Altered versions of the well-established energy detection, maximum eigenvalue detection, maximum-minimum eigenvalue detection and generalized likelihood ratio test techniques are considered. In what concerns the fusion rules, we consider the sample fusion, the decision fusion and a recently proposed eigenvalue fusion technique. It is shown that, in spite of the intrinsic interference present in the continuous mode, this approach is flexible in terms of the sensing time, eventually allowing for better sensing performance.