Research On Spectrum Allocation Model And Algorithm In Cognitive Radio Network

With the progress of science and technology, and the rapid development of human society, new wireless communication businesses and equipments are constantly increasing. The problem of shortage of radio spectrum resources is becoming increasingly prominent. However, the spectrum allocation in a static and fixed mode leads utilization of spectrum resources to be very low and very unbalanced. Therefore, wireless spectrum resource is facing a contradiction between "shortage" and "waste". In this background, cognitive radio technology viewed as a new communication technology utilizing spectrum resource efficiently emerges as the times require and it becomes one of the hottest technologies in the field of wireless communication. This paper mainly researches on spectrum allocation in cognitive radio and includes the following aspects:Aiming at the problem of existing evolutionary algorithm applied on the spectrum allocation based on graph coloring mode in low convergence accuracy, this paper design a quantum immune clone algorithm based spectrum allocation approach in cognitive radio. This approach takes advantage of the ability of global search and rapid convergence of quantum immune algorithm to search optimization. Considering updated solution can not the solution of constrained problem directly and aiming at assignment targets of fairness and bottleneck reward, this paper designs a fairness constraint operator and a bottleneck reward constraint operator. These two operators ensure the population meet the constraint and improves searching ability of the algorithm. The simulation results show that the two operators have effect on the condition of limited spectrum resource and serious spectrum interferences. Compared with the reference of optimization algorithm, the proposed approach has better searching ability and fast searching speed.We further research on spectrum allocation in cognitive radio. A method expressing the cognitive users, available bandwidths and the relationships between users in a unified way is proposed by constructing an undirected graph model. Then the spectrum assignment problem is converted into optimizing the target function of undirected graph, which guarantees to consider both the benefit of cognitive users and the interference between them at the same time. The corresponding optimization procedure is implemented through designing an expanded iterative conditional mode algorithm, which can realize high precision local search and global optimization. So, it can improve the chances to find the optimal solution. Experiment results show that the proposed method can assign the spectrum effectively and perform well on different networking environments. It has more advantage when the network scale is larger and spectrum conflict is worse. Meanwhile, it is less affected by the number of population which reflects the stability of the proposed method.