利用遗传算法改进的路由和波长分配算法

Routing and wavelength assignment algorithms improved by genetic algorithms

给定拓扑结构和流量请求的波分复用静态波长路由光网络中 ,利用路由和波长分配算法求得的网络最小所需波长数与网络节点编号有关。为了得到网络最小所需波长数所对应的网络节点编号 ,首次利用遗传算法对节点编号进行优化。和已有的统计方法相比较 ,采用遗传算法可以有效地消除原有方法的随机性和盲目性。模拟计算表明 ,利用遗传算法可以更快地求得网络最小所需波长数所对应的网络节点编号。以此求得的网络所需波长数 。

In a wavelength division multiplexing (WDM) static wavelength routed optical network with given network topology and traffic load, the minimum number of required wavelengths obtained by the routing and wavelength assignment algorithms is associated with the node numbering order in the network. The optimum node numbering order corresponding to the minimum number of wavelengths required can be found with a genetic algorithm. The genetic algorithm can eliminate the stochastic behavior and blindness of the original statistical method. Numerical simulation shows that the optimum node numbering order is more quickly found by the genetic algorithm, and the minimum number of wavelengths required corresponding to the node numbering order is closer to the lower bound.