A Novel Signal Separation Algorithm for Wideband Spectrum Sensing in Cognitive Networks

In cognitive radio networks, since any cognitive terminal can access the wideband frequency spectrum, interference from hostile terminals can be hard to detect and eliminate. Since such interference are supposedly wideband compared to other signals, it would be difficult to avoid them. In order to handle hostile terminal interference, signal spectrum sensing and separation is a promising approach that allows the application of proper strategies against hostile terminals. In this paper, we propose a novel signal separation algorithm for spectrum sensing and signal separation, which locates and separates signals occupying the wideband frequency spectrum. The proposed algorithm consists of two steps: 1) wavelet edge detection algorithm is adopted to locate the signal spectrum edge; 2) wideband signal separation approach deals with signal separation and recovery. The analytical results indicate that this novel spectrum sensing algorithm is able to separate wideband signals accurately. Simulation results show that occupying signals can be successfully separated and recovered from wideband spectrum.