摘要
为满足用户新的应用需求或者添加额外的安全保护功能,研究了基于网络功能虚拟化的云网络中已部署网络服务实例的功能扩展问题。将云网络中的网络服务扩展建模为整数线性规划模型,考虑到初始服务部署不受影响、扩展图部署、资源容量和节点亲和性等约束。同时,为应对未来大规模云网络的动态性和复杂性,提出了一种基于量子遗传算法的网络服务扩展算法,该算法基于量子态通过量子并行运算来提高求解效率和求解质量。仿真结果表明,所提算法在扩展成功率和平均资源开销等方面均表现较好,且具有较低的时间复杂度。
To meet the client’s new business requirements or add additional security protection functions,the already hosted network service extension problem in the cloud network based on network function virtualization was researched.The network service extension in the cloud network was modeled as an integer linear programming,considering the im-pact on initial service,extended graph deployment,resource capacity and virtual network function affinity constraints,and so on.To deal with the computational complexity and dynamism of future large-scale cloud networks,a QGA-based network service extension algorithm was proposed to improve solution efficiency and solution quality through quantum parallel computing.Simulation results prove that the efficient performance of the proposed algorithm in terms of exten-sion successful ratio and average resource costs,and the proposed algorithm has low time complexity.
作者
邱航
汤红波
游伟
赵宇
柏溢
QIU Hang;TANG Hongbo;YOU Wei;ZHAO Yu;BAI Yi(Institute of Information Technology,Information Engineering University,Zhengzhou 450002,China)
出处
《通信学报》
EI
CSCD
北大核心
2022年第11期44-52,共9页
Journal on Communications
基金
国家重点研发计划基金资助项目(No.2020YFB1806607)。
关键词
网络功能虚拟化
网络服务扩展
虚拟网络功能转发图
量子遗传算法
network function virtualization
network service extension
VNF-forwarding graph
quantum genetic algorithm
