摘要
将免疫系统的免疫机制引入到粒子群优化算法的设计中,模拟免疫系统、群集智能和神经网络的信息处理机制,提出了免疫粒子群优化算法。这种免疫粒子群算法结合了粒子群的近似全局优化能力和由Hopfield神经网络构成的免疫系统的快速信息处理机制,加快了算法的收敛速度,并提高了粒子群算法的全局收敛能力。然后利用此算法对CDMA系统的多用户检测性能改进问题进行实验研究,证明了本文的方法有较快的收敛速度,并且无论是抗多址干扰能力还是抗远近效应能力都优于传统方法和一些应用优化算法的多用户检测器。
In this paper,idea on immune information processing mechanism of immune system is introduced to a particle swarm optimization algorithm.By simulating the information processing mechanism of biological immune system and particle swarm intelligence,we present an hnmune Particle Swarm Optimization algorithm （IPSO）.The proposed algorithm is a hybridization of the PSO that has better performance and an immune operator that reduces the computational complexity by providing faster convergence with Hopfield neural network.Experiments that have been carried out on the algorithm can improve the performance of muhiuser detection in CDMA systems.Simulation results show that IPSO detector not only can achieve the global optimization in fast convergenee rate,but also is obviously superior to the conventional receiver and the muhiuser detectors based on previous optimum algorithms in terms of the muhiple-aecess interfere and near-far resistance.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第35期148-151,共4页
Computer Engineering and Applications
基金
哈尔滨市科学研究基金(2005AFXXJ033)。
