摘要
随着网络规模的不断扩大,一些基于纯数学模型的路由算法已经面临新的挑战。针对网络路由对实时性和可靠性的要求,采用动态路径诱导算法对网络流量进行实时预测,可以解决传统的诱导算法存在的时变性差和收敛慢的问题。动态路径诱导算法基于神经网络时间预测模型和遗传算法,经仿真表明该动态路径诱导方法在网络繁忙时能显著改善网络路由性能。
With the constant expansion of the network size, some routing algorithms based on pure mathematical models have been confronted with new challenges. In order to meet the requirements for real-time and reliability of network routing, a new dynamic route guidance method resolved the limitation of traditional dynamic route guidance algorithm by forecasting the network traffic and composing real-time road weight matrix. This method is based on Neural Network (NN) and Genetic Algorithm (GA), and it has been proven by lab experiments that it can significantly optimize the performance of network routing in the busy network.
出处
《中兴通讯技术》
2007年第6期48-50,共3页
ZTE Technology Journal
关键词
遗传算法
动态路径诱导
最优路径
神经网络
genetic algorithm
dynamic route guidance
optimal route
neural network