改进遗传算法的光网络波长分配研究

Research on wavelength assignment in optical networks based on improved genetic algorithm

常规方法计算波长分配目标函数时,忽略了波长利用率的优化,导致光网络利用率较低、阻塞率较高,为此,提出了基于改进遗传算法的光网络波长分配方法。以最小化调度时间、波长利用率,最大化光路链路质量、任务请求建立比例为目标,计算波长分配目标函数,根据业务请求优先级的首选波长集,获得光网络可用波长,遵循波长分配约束条件,搜索可用波长中的可分配波长,形成改进遗传算法初始种群,将目标函数作为算法适应度函数,输出种群最优染色体,得到路由链路和波长的最优匹配结果。将高速信息示范网作为仿真对象,仿真光网络拓扑结构,实验结果表明,面对不同的业务数量和业务强度,设计方法相比常规方法,提高了光网络利用率,降低了阻塞率,有效改善了光网络性能。

When calculating the wavelength assignment objective function,the conventional method ignores the optimization of wavelength utilization,resulting in low utilization and high blocking rate of optical network.Therefore,an optical network wavelength assignment method based on improved genetic algorithm is proposed.In order to minimize the scheduling time and the wavelength utilization,maximize the optical link quality and the establishment proportion of task requests,the wavelength allocation objective function is calculated,the available wavelengths of the optical network is obtained according to preferred wavelength set of service request priority,the wavelength allocation constraints is followed,the assignable wavelengths in the available wavelengths are searched,and the initial population of the improved genetic algorithm is formed.Taking the objective function as fitness function of the algorithm,the optimal chromosome of the population is output,and the optimal matching results of routing link and wavelength are obtained.The high-speed information demonstration network is taken as the simulation object to simulate the optical network topology.The experimental results show that in the face of different service quantity and intensity,the design method improves the optical network utilization,reduces the blocking rate and effectively improves the optical network performance compared with conventional method.