波长转换是应用在全光WDM网络中打破波长一致性约束、降低网络阻塞概率的一种有效技术,但限于波长转换器的高昂成本以及其对信号失真度的影响,目前还不可能为网络中的每个节点均配置波长转换器,所以波长转换器需要以最优方案配置在网络中的个别关键节点上,本文提出了一种在任意拓扑结构的网络中,基于遗传算法的波长转换器配置算法,并对该算法在NSFNET网络上进行了计算机仿真,模拟结果显示该算法非常有效.
In this paper, the optimal placement of wavelength converter problem is studied. Because of less iterations and fast convergence, genetic algorithm is used for placing full range wavelength converters in arbitrary WDM networks. Simulations have been carried out on 14-node NSFNET, the result shows that the algorithm is very efficient, and optimal solutions can be achieved.