一种改进遗传算法在WDM光网络中的应用  

Utilization of an enhanced genetic algorithm in WDM optical networks

作　　者：何军[1] 李勇[2] 李蔚[1] 刘德明[1] 

机构地区：[1]华中科技大学光电子工程系,湖北武汉430074 [2]武汉邮电科学研究院,湖北武汉430074

出　　处：《光通信研究》2004年第5期15-16,35,共3页Study on Optical Communications

基　　金：武汉市科技攻关基金资助项目(2002100513004)

摘　　要：在波长路由光网(WRON)中,波长转换对提高网络性能、减少阻塞率起着重要的作用.但由于目前全光波长转换器非常昂贵,因而如何最优利用波长转换是配置光网络的一个重要问题.文中利用遗传算法很好地解决了这一问题,并且利用数值模拟的方法,计算了美国自然科学基金网(NSFnet)网络中最优配置下网络的阻塞情况.然后针对普通遗传算法的缺点,提出了一种改进遗传算法.最后利用数值模拟的方法,比较了两种算法的性能.In a wavelength routing optical networks (WRON), wavelength converters can greatly enhance the fiber utilization and reduce the overall call blocking probability of the networks. But wavelength converters are very expensive nowadays. So the optimal utilization of wavelength converters is an important question in optical network configuration. This paper resolved this question by using genetic algorithm (GA). We calculated the blocking probability of optimal configuration in NSFNet with simulation. Considering normal genetic algorithm's shortcomings, we proposed an enhanced genetic algorithm (EGA). At last two GAs's performances were compared by simulation.

关 键 词：波分复用光网络 光波长转换 遗传算法 阻塞率 

分 类 号：TN915[电子电信—通信与信息系统]