基于混合遗传算法的虚拟网络映射方法研究  

Research on Virtual Network Mapping Based on Mixed Genetic Algorithm

作　　者：刘佳[1] 宋涛[1] 胡颖[1] 庄雷[1] 

机构地区：[1]郑州大学信息工程学院,郑州450000

出　　处：《小型微型计算机系统》2016年第4期773-777,共5页Journal of Chinese Computer Systems

基　　金：国家"九七三"重点基础研究发展计划项目(2012CB315901)资助;河南省科技厅攻关项目(122102210042)资助;国家自然科学基金项目(61379079)资助

摘　　要：虚拟网络映射是网络虚拟化研究的关键内容,利用传统遗传算法解决虚拟网络映射问题,由于遗传算法本身的缺点使得问题容易过早进入局部最优解,且收敛速度慢.在基本遗传算法中加入改进的单纯形算法,以最大化In Ps的收益为目标,建立混合整数线性规划(MILP)模型,提出VNE-M-GA的虚拟网络映射算法.该算法利用单纯形法预估寻优方向,遗传算法和单纯形法迭代优化映射方案,尽可能的避免局部最优.实验结果表明该方法解决虚拟网络映射问题,与现有算法实验结果相比,一定程度改进了早熟收敛问题,提高了In Ps总收益与虚拟网络请求接受率.Virtual Network Embedding problem is very important to the future development of the network,using the basic genetic algorithm to solve the problem of the virtual network embedding,due to the limitations of genetic algorithm makes the problem easy to early into the local optimal solution,and the slowconvergence speed.Regarding VNE as an M ixed Integer Linear Programming model with a goal to maximize revenue,a newVNE algorithm based on mixed genetic algorithm was proposed.The simplex method is used to estimate the optimal direction,and the proposed algorithm took advantage of genetic algorithm and simplex method to optimize mapping scheme,as much as possible to avoid local optimal.Compared with the existing approaches,the experimental results demonstrate that the proposed algorithm can increase the revenue of In Ps and the acceptance ration of virtual network requests.

关 键 词：虚拟网络映射 混合整数规划 遗传算法 单纯形法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]