Parallel Architecture Implementation of Spectrum Sensing Based on Covariance in Cognitive Radio

Spectrum sensing is one of the most important technologies in cognitive radio communication system. Considered the different characteristics of spectrum sensing algorithms and the computational complexity of the implementation in reality, we propose a method to process the sample signals with parallel architecture for spectrum sensing based on statistical covariance matrix. Through reasonable data grouping without affecting the performance of detection, it can take advantage of parallel computing in hardware to spectrum sensing rapidly and real-time. This parallel structure plays a great application values to reduce the complexity when implementation. This implementation method has been applied in the test bed of cognitive radio communication system which worked on the spectrum from 694MHz to 806MHz. Finally, the verification of this method has been done on the demo system based on DVB-T signal.