基于遗传粒子群的认知无线电频谱分配  

Spectrum Allocation Based on Genetic Particle Swarm Algorithm in Cognitive Radio System

作　　者：林培培[1] 杨铁军[1] 

机构地区：[1]河南工业大学信息科学与工程学院,河南郑州450001

出　　处：《无线电工程》2014年第9期12-15,共4页Radio Engineering

基　　金：国家自然科学基金资助项目(102102210121);河南工业大学研究生教育创新计划基金资助项目(2012YJCX55)

摘　　要：针对传统遗传算法和粒子群算法在实际应用中存在早熟和收敛效率低的问题,引入了遗传算法和粒子群算法的混合算法。在混合算法中引入种群分割策略,避免了算法陷入局部最优解;引入粒子群重构变异算子避免了算法的早熟,并提高了算法的收敛效率。针对系统效益和时间开销等性能指标进行了仿真,仿真结果表明,改进算法与传统遗传算法和粒子群算法相比,能够提高搜索效率,得到更好的系统效益。The traditional genetic algorithm and particle swarm algorithm have problems of prematurity and low convergence efficiency in the practical application. In order to solve the problems,an algorithm based on a hybrid of genetic algorithm and particle swarm algorithm is introduced. The population segmentation strategy is introduced to avoid algorithm trapped in local optimal solution. The particle swarm refactoring mutation operator is introduced to avoid prematurity and improve the convergence efficiency of the algorithm. Finally,performance indicators of time cost and system efficiency are simulated. The improved algorithm is compared with traditional genetic algorithm and particle swarm algorithm. Simulation results show that it performs better both in terms of convergence speed and system efficiency.

关 键 词：认知无线电 频谱分配 图论着色 遗传算法 粒子群算法 

分 类 号：TN92[电子电信—通信与信息系统]