Compressive Spectrum Sensing Based on Sparse Sub-band Basis in Wireless Sensor Network

An approach based on Sparse Sub-band Basis (SSB) for compressive spectrum sensing in Wireless Sensor Network (WSN) is presented in this paper, considering the unsatisfactory accuracy and complex calculation of the traditional ones. It is proved that the SSB matches not only the orthogonality and completeness of the basis, but also the Restricted Isometry Property (RIP) in reconstructing the signal. The simulation results show that the reconstruction based on SSB can detect the accurate location and amplitude of spectrum occupancy, and is more robust than traditional edge detection method. Additionally, this approach has higher compression ratio and less calculation, which is suitable for nodes in WSN.