一种基于改进神经网络的动态信道分配算法

提出了一种基于改进神经网络的整体优化的动态信道分配方案(GODCA),该算法尽量保证最大程度的紧致分配,网络优化采用Hopfield神经网络,克服了传统动态信道分配方案实现方式繁杂,且往往不能达到紧致分配方式的不足,同时针对神经网络算法随机性强的弱点,提出一种优化的收敛算法。分析和仿真表明,该算法不论在业务量较大,还是较小都有较小的呼阻率和较高的频谱利用率。

A Dynamic Channel Assignment Algorithm Based on Improved Neural Network

The paper presents a Globally Optimized Dynamic Channel Assignment(GODCA) scheme based on improved Neural Network for cellular mobile communications systems.It differs from other strategies by consistently keeping the system in the utmost optimal state.It employs the Hopfiled neural network for optimization,which avoids the complicated assessment of channel compactness and guarantees optimum solution for every assignment.Directed at the weak capacity of convergence of genetic algorithm,an optimal algorithm is constructed.The analysis and simulations show that the algorithm can attain low blocking probability disregarding the traffic density of the system,and GODCA is shown to be superior to the current FCA and DCA in terms of call blocking probability and spectrum utilization.