Geolocation-based Real Time In-band Spectrum Sensing for Cognitive Networks

摘要

The widely used "quiet periods" have at least two shortcomings: periodical interrupting the secondary transmission will degrade the QoS of secondary networks, e.g., decreasing the system capacity, increasing end-to-end packet delay; tight synchronization, which is difficult to be implemented in the distributed networks, becomes necessary. Aiming to overcome such shortcomings, a geolocation based real time in-band spectrum sensing is proposed in this paper. By using the locations of the primary transmitters and the secondary users, the proposed method is able to differentiate the target primary signal power from the co-channel and adjacent-channel interference produced by the simultaneous data transmissions in the secondary networks. Consequently, the interference can be cancelled when constructing the test statistic for the energy detectors. No prior knowledge of the primary signal is required by the proposed method. Both analyzing and simulation results show that the proposed method is able to successfully cancel the interference and can achieve a good performance.