Study Of Spectrum Sensing Algortihms For Cognitive Radio Systems

This paper has studied on spectrum sensing technology, one of the key technologies of cognitive radio systems, and its application of wireless body area networks.After the statement of the basic conceptions of cognitive radio system, system model, related standards, the research achievement around the world, this paper introduced the basic theory of spectrum sensing and the way of estimating the performance of spectrum sensing. Three main spectrum sensing technologies are introduced, the theory of cyclostationary feature detection is particularly described. The advantages and disadvantages of each technology are discussed and compared. In addition, the cooperative sensing technology is also described.This paper stressed on the researches, analyses and simulations of cyclostationary feature detection. The frequency smoothing estimation method for cyclic spectrum in discrete domain is introduced as the precondition for cycle detection. The disadvantage of cyclostationary feature detection is described and the possible ways to improve it are suggested in this paper. Two simplified cyclic based detection method are given in this paper. The first one is the cycle frequency based detection which calculates the spectrum correlation density of the cycle frequencies to complete the spectrum sensing, the second one is the zero-frequency based detection which calculate the spectrum correlation density of the frequency equals zero, and the extra advantage of this method is that it doesn't need any prior knowledge for application. Apart from that, this paper also discusses cyclic spectrum analyses of signal with additive noise, the detail simulation results are given in this paper. And the performance is compared with the energy detection method to illustrate the benefit of the cycle detection. In order to further analysis of the cyclostationary feature detection method, the cycle spectrum noise model is discussed, the extreme value theory is describe in this paper. After the simulation and study of the AWGN, the noise of cycle spectrum can be best describe by the first type model of General Extreme Value Distribution. And the cycle spectrum noise is made a statistic analysis, the fitting lines of the data distribution is given in the simulation results.At last in this paper, the concept of wireless body area network (WBAN) is discussed, the detection method proposed in this paper are further simulated in the WBAN environment. And the conclusion will not only list some research gains but also will give advice for later research in this field.