Cooperative Spectrum Sensing Based On Multi-nodes In Cognitive Radio Networks

With the rapid growth of wireless communications, spectrum scarcity becomes more and more serious under the low utilization of spectrum resources. Facing with the conflict, cognitive radio emerges as the solution of spectrum scarcity. It is an intelligent wireless telecommunication system allowing efficient utilization by spectrum sensing, self adaptation and dynamic spectrum sharing according to its environment sensing. Therefore, cognitive radio is recognized as an agile, intelligent and reliable communication system with ability of occupying plenty of unused spectrum under the condition that it don't cause interference to licensed system. Spectrum sensing is the core technology of cognitive radio, Cognitive user senses the environment to find the"spectrum hole", thus increasing the frequency spectrum utilization without affecting the licensed user. Because of the effect of multipath effect and the deep fading,"hidden terminal"problem has severely affected the performance of spectrum testing. Collaborative spectrum sensing can solve the hidden terminal problem and increase the system sensing performance. From several aspects, this paper focus on collaboration spectrum testing which is based on node and researches parameter optimization on the system sensing costs and system throughput. This thesis also puts forward the collaborative detection algorithm based on clustered network.This thesis studies spectrum sensing based on single node. Spectrum sensing technology are comprehensively introduced, including three methods which is respectively based on matched filter,cyclostationary characteristics and energy detections. We focus on the method based on energy detection and its system model. The energy detection is simulated ,and the performance analysis of system sensing is given with curve describing detection probability,alarm probability,SNR and sampling points. We also research cooperative spectrum sensing based on multi nodes and analyse the system model. The survey of cooperative spectrum sensing using different data fusion rules is given. MATLAB simulation is presented, comparing every algorithm performance.The more users engaged in cooperative spectrum detection, the better the system performance of spectrum detection. However, when more users collaborate, detection overhead increases rapidly. We put forward the cooperative spectrum detection based on optimal number of users to solve the problem how to optimize the number of users in cooperative spectrum detection under the condition that the system resource is limited. The thesis analyses the optimal number of users problem in cooperative spectrum detection which is respectively based on"and"and"or"data fusion rule. The performance of problem in Rayleigh and Nakagami fading channel is stimulated in order to find the optimal number.The research on cooperative spectrum sensing based on 802.22 protocol is viewed from optimal system throughout. We introduce the concept and system model of 802.22 protocol. The system throughout based on cooperative spectrum sensing is proposed with the parameters of spectrum sensing. And mathematical formula is presented. In this thesis, we study the optimization of parameters based on system throughout. The mathematical deduction proves that there exists certain false alarm probability and subscribers which makes system throughout optimal. The mathematical formula is deduced.In cognitive radio we use hierarchical networks. Users are divides into clusters:Common cluster member and cluster head constitute the first level of network; clusters make the higher level networks, which is used for long distance transmission. Hierarchical networks can save much energy while it ensures the range of data communication. In the paper, we put forward a improvement algorithm, which combines multi users diversity in cooperative spectrum and dynamic clustering in LEACH protocol. In the cluster voting process, hierarchical voting is used to enhance the system sensing performance. Simulation results show that the new algorithm has good performance. It enhances system sensing capability, ensuring the whole network's energy use ratio.