论文部分内容阅读
采用模糊逻辑和神经网络技术进行异构无线网络接入选择的方法未合理考虑网络负载状况,为此提出一种对网络负载具有很好动态适应性的基于粒子群优化(PSO)模糊神经元的接入选择方法.该方法将可接入网络的接入阻塞率相等作为模糊神经元参数学习的目标,并结合具有全局寻优能力的PSO算法设定参数初值,提高了参数学习精度.仿真结果表明,该方法能有效实现网络间负载均衡,相对于最大负载均衡算法可降低网络的接入阻塞率.
The method of using fuzzy logic and neural network for heterogeneous wireless network access selection does not reasonably consider the network load condition. Therefore, a particle swarm optimization (PSO) fuzzy neuron with good dynamic adaptability to network load is proposed This method takes the access blocking rate of accessible network as the goal of fuzzy neuron parameter learning and combines the initial parameters of PSO algorithm with global optimization ability to improve the parameter learning precision. The results show that this method can effectively achieve load balancing among networks, and reduce the access blocking rate of the network relative to the maximum load balancing algorithm.