Abstract
In this study, the authors propose a secure cooperative spectrum sensing algorithm that is based on compressive sampling theory and employed in wide-band cognitive radio network. The proposed algorithm seeks to reduce the hardware complexity of the spectrum sensing node by reducing the reported data from this node to the fusion centre and/or other sensing nodes. Due to the spatial correlation between the received primary users' signals at the sensing nodes, the difference between the measurement vectors of any two sensing nodes is expected to be very small when it is compared with the value of the measurement vector itself. The expected reduction in the dynamic range of the difference measurement vector motivates to use the 1-bit quantiser where only 10bit per measurement is retained, representing its sign. Moreover, the proposed algorithm is impregnable against spectrum sensing data falsification attacks. Both the fusion centre and sensing nodes utilise the proposed authentication algorithm to pick out data from the trustful nodes only to estimate its output data. According to the simulation results, the proposed system guarantees more safeguard to the license primary user while preserves the unlicensed secondary user throughput.