基于免疫遗传算法的多用户检测器设计

Design of Multi-user Detector Based on Immune Genetic Algorithm

文中将免疫系统的免疫机制引入到遗传算法的设计中,模拟免疫系统、神经网络、遗传系统的信息处理机制,提出了免疫遗传算法。这种免疫遗传算法结合了遗传算法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,并提高了遗传算法的全局收敛能力。然后在在CDMA系统利用此算法设计了多用户检测器,仿真结果证明了本文的方法能够快速收敛到全局最优解,并且无论抗多址干扰能力和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 genetic al- gorithm.By simulating the information processing mechanism of biological immune system and genetic system and neu- ral network.We proposed the immune genetic algorithm(IGA).The proposed algorithms have both the properties of the original genetic algorithm and the faster convergence of immune system which is consist of neuron of Hopfield neural network,can improve the abilities of seeking the global excellent result and convergence speed.Experiments have been carried out on the method can improved on the multiuser detection in CDMA systems,the simulation results show that IGA detector can achieve the global optimization in fast convergence rate,and also show that the proposed detector is obviously superior to the conventional receiver and multiuser detectors based on some optimum algorithm in terms of the multiple access interfere and the mitigation of near-far effect.