遗传算法在光网络动态流量重构疏导中的应用

Application of GA in Reconfiguration Grooming Problems of Dynamic Traffic in Optical Networks

针对当前网络的动态业务,提出动态业务流量的重构疏导方案,并给出改进的遗传算法.此算法能通过调整网络的逻辑或物理拓扑结构以适应各种动态变化的业务.实验结果表明:改进的算法能搜索到较优的结果,即添加较少的ADM和波长就可使网络继续承载所有新、旧动态业务.但是算法复杂度较高,因而消耗的时间较多.

To deal with dynamic traffic in present networks, this paper addresses the reconfigurable grooming of dynamic traffic and proposed an improved genetic algorithm. The algorithm can make the network fit all kinds of dynamic traffic through dynamically adjusting the virtual or physical topology of the network. Simulations show that the algorithm, though comparatively complicated and timeconsuming, can achieve a satisfying solution.