A Study Of Spectrum Sensing Technology On Cognitive Radio

In recent years, the rapid developing wireless communication technology and the growing various wireless devices demand more and more wireless spectrum resources which are not enough available. The study found that the phenomenon of spectrum holes is very serious due to the bad policy of allocation, and the use of spectrum is also not balanced. Therefore, as a key technology of cognitive radio system, spectrum detection technology becomes more and more important.In this paper, the concept and the critical technologies of cognitive radios are firstly introduced, and then the traditional spectrum detection method are analyzed including single-point spectrum sensing method and collaborative spectrum sensing method. The performance of these methods is analyzed. The Single-point spectrum detection algorithm based on the transmission side includes three methods: matched filtering method, energy detection method and cyclostationary feature detection method. Each of them have advantages and disadvantages: The matched filtering method needs short time to detect, but it would require priori information of primary users; Energy detection method is simple but it needs a long time to detect and cannot distinguish the signal type; Cyclostationary feature detection method can distinguish the signal type, but it is complex. Cooperative spectrum detection includes hard decision and soft decision. Hard decision have three methods including AND, OR and Majority Fusion criterion. In this paper, all the three methods are simulated and analyzed for their performance. In addition, a new cognitive sensing spectrum detection algorithm based on compressed sensing is introduced. Firstly, the fundamental of compressed sensing is given, then the mathematical basis is analyzed, and we combine the compressed sensing technology with cognitive radio spectrum sensing technology together and draw the MMP algorithm based on cooperative spectrum sensing.This method makes a full use of the coefficient characteristic of channel with incomplete information. In the traditional collaborate spectrum sensing method, the fusion center needs to determine the ideal channel position and make a judgment on a large number of perceptual information. While the large amount of information is the bottleneck of the collaborative spectrum sensing applications. In contrast, the MMP algorithm based on compressed sensing reduces the required perceptual information greatly and can achieve good detection performance. At last, the cognitive radio spectrum detection technology is applied in the wireless body area network. With the energy detection algorithm and cyclostationary feature detection technology combined, the performance of wireless body area network is improved.