差分头脑风暴算法及其在频谱感知中的应用

Differential brain storm optimization algorithm and its application to spectrum sensing

为了有效求解连续优化问题,基于差分进化算法和头脑风暴优化算法的智能演进原理,提出一种新的全局搜索算法,即差分头脑风暴算法。通过4个经典的基准函数对该算法进行测试,并将该算法应用于频谱感知这个认知无线电领域的热点问题,提出基于差分头脑风暴的协作式频谱感知算法。使用差分头脑风暴算法、头脑风暴算法、混合蛙跳算法以及粒子群算法进行仿真对比。仿真结果表明,所提出的算法基于设计的创新方程,具有很强的全局收敛能力,能够显著改进头脑风暴算法的性能;基于差分头脑风暴的频谱感知检测概率比其他算法都高,且收敛速度比头脑风暴算法提高至少3倍。

In order to solve continuous optimization problems effectively,a novel global search method named differential brain storm optimization（ DBSO） algorithm was proposed based on the intelligent evolutionary principles of brain storm optimization（ BSO） algorithm and differential evolution（ DE） algorithm. The proposed algorithm was tested with four classic benchmark functions,and applied to hot spectrum sensing issue in the domain of cognitive radio,and a cooperative spectrum sensing method was proposed based on DBSO algorithm. Comparison was conducted between DBSO algorithm,BSO algorithm,shuffled frog leaping algorithm（ SFLA） and particle swarm optimization（ PSO） algorithm. The simulation results of the benchmark functions show that in terms of convergence precision,the presented DBSO algorithm based on new creating equations outperforms other 3 intelligent algorithms;compared with these three intelligent spectrum sensing algorithms,the detection probability of DBSO is higher,and the convergence speed of the DBSO increases 3 times compared to BSO algorithm.