The Research Of Spectrum Sensing In Cognitive Radio System

Spectrum scarcity is the fundamental problem for wireless communication systems. To solve this problem, researchers tried to develop more sophisticated wireless communication technologies. Existing static spectrum allocation policy only allow communication systems to work in the spectrum bands which are allocated to them, however, practical measurements show that the usage of the allocated spectrum is very low. This suggests that existing static spectrum allocation policy wastes many spectrum resources and that dynamic spectrum access technologies can greatly enhance spectrum efficiency, and then solve the spectrum scarcity problem.Cognitive Radio (CR) is a promising technology to enable dynamic spectrum access. At first, CR finds idle spectrum bands via spectrum sensing, then, CR uses dynamic spectrum access and allocation technologies to allocate the idle spectrum bands. The spectrum sensing is a key technology in CR. Efficient spectrum sensing can not only avoid interference to existing systems, but also effectively find idle channels to increase spectrum efficiency. This thesis focuses on the research of spectrum sensing technologies, and the main innovations and contributions are listed as follows:1. Proposed novel energy detection based on Bayesian Estimation. The energy detection is a promising spectrum sensing technology in CR; however, it performs poorly when the noise power can not be estimated accurately. The novel energy detection does not need the knowledge of noise power, and thus can completely eliminate the detrimental effects of inaccurate noise power estimation. Moreover, the novel energy detection has the same performance as the conventional energy detection when the noise power can be accurately estimated. 2. Derive the optimal cooperative spectrum sensing method to achieve the miximum channel capacity. Most of existing researches of spectrum sensing are based on the Neyman-Pearson criterion, which aims to maximize the probability of detection given the probability of false alarm. However, this criterion does not relate the spectrum sensing with the achieved channel capacity. Our optimal sensing method can set the sensing parameters to maximize the channel capacity.3. Under the regime of low SNR, derive the quantitative relation between the hard and soft cooperative spectrum sensing methods: when the number of sensors in hard cooperation is 1.6 times as that of soft cooperation, they have nealy the same performance. In hard cooperation, the sensors at first perform individual spectrum sensing, and then send their decisions to a fusion center for final decision. In soft cooperation, the sensors send their measurements to the fusion center for the final decision. Our conclusion can help the selection of cooperation type in spectrum sensing in practical systems.