Research On Parameter Optimization And Sensitivity Analysis In Cognitive Radio Networks

Cognitive radio technology is regarded as an intelligent wireless technology which can reuse spectrum in a secondary way and improve spectrum utilization significantly, and provides a new way for dynamic spectrum sharing.Relay on the NSF project which named 'The topology management, MAC mechanism and adaptive resource allocation of Cog-MESH network', this paper focuses on the research of parameter optimization and sensitivity analysis in cognitive radio systems. Parameter optimization of cognitive radio can real-time obtain current optimal operating parameters according to changes of outside environment; yet parameter sensitivity analysis is to seek the internal relation between transmission parameters and performance objectives of systems. Combined parameter sensitivity analysis with parameter optimization, which can both guide adaptive adjustment of transmission parameters, and can shoot at the target in paying close attention to major factors and ignoring secondary causes, thereby reduce the complexity and computation of signal processing. The contents include:1) Started from the basic principles and crucial technologies of cognitive radio, analyzed and compared typical cognitive radio decision engines and technical characteristics of cognitive engine models, and described the mathematical description of multi-objective optimization and traditional multi-objective optimization methods.2) Cognitive radio can adaptively adjust its working parameters according to users'needs and change in the environment. Most of the existing cognitive engines use genetic algorithm to optimize parameters, however with the increase in the number of cognitive users, the increased chromosome results in long convergence time of genetic algorithm, which can not meet the needs of real-time communication. Therefore, when optimizing cognitive radio parameters, seeking an optimization algorithm which is high efficiency, fast convergence and high stability, is one of the important research topics of cognitive radio. Based on particle swarm algorithm, according to transmission parameters of cognitive radio, optimization goals and their fitness functions, an improved inertia factor particle swarm optimization which is used for parameter optimization in cognitive radio is proposed, furthermore simulation proves that this algorithm works well, and has better convergence, efficiency than genetic algorithm, stability of this algorithm is higher, and can meet real-time processing requirement of cognitive radio. 3) Parameter sensitivity analysis is to obtain and quantize the influence degree of every transmission parameters to cognitive engine decision process. Combined parameter sensitivity analysis with parameter optimization, parameters sensitivity analysis is done on transmission parameters of cognitive radio in different communication modes respectively, and we can selectively remove lower sensitivity parameters from the objective function, in order to reduce system complexity and processing delay.