免疫遗传算法及认知无线电参数优化决策.pdfVIP

2012年第02期，第45卷 通 信 技 术 Vol.45，No.02,2012 总第242期 Communications Technology No.242,Totally 免疫遗传算法及认知无线电参数优化决策 周德全 （浙江农林大学信息工程学院，浙江 临安 311300 ） 【摘 要】根据感知的频谱环境变化及时优化并调整无线电参数是认知无线电的关键技术之一，也是一 个复杂的非线性多目标优化决策问题。遗传算法是最适合优化问题的，但当遗传算法应用于优化问题时存 在过早收敛问题。提出了基于遗传算法和人工免疫系统相结合的免疫遗传算法（IGA）来克服以上问题。由 于在GA算法中引入了免疫系统中抗体和抗原的概念并在每一次迭代中丢弃亲和力较大的抗体，有效地防止 了GA中过早收敛现象。最后，用免疫遗传算法来解决认知无线电的参数优化问题。仿真结果表明，免疫遗 传算法可以迅速达到最优决策。 【关键词】遗传算法；认知无线电；多目标优化；人工免疫系统 【中图分类号】TN914; TP182 【文献标识码】A 【文章编号】1002-0802(2012)02-0053-03 Immune Genetic Algorithm and Parameters Optimization of Cognitive Radios ZHOU De-quan （School of Information Sci. and Tech., Zhejiang AF University, Hangzhou Zhejiang 311300, China） 【Abstract】Timely optimization and regulation of the parameters in CR system according to spectrum sensing is one of the CR key techniques, and also a multi-objective optimization problem. Genetic algorithms are best suitable for optimization problem. However, premature convergence exists when genetic algorithm is applied to optimization problem. The immune genetic algorithm (IGA) in combination of artificial immune system and GA together is proposed, thus to overcome this problem. Due to the introduction of some antibodies, IGA could effectively prevent iteration from premature convergence, while the antibodies with much greater affinity in each iteration are rejected. Finally, the IGA is used to solve parameter optimization problems of cognitive radios. Simulation indicates that IGA can rapidly reach an optimal decision. 【Key words】genetic algorithms; cognitive radios; multi-objective optimization; arti