基于流量优化的可靠服务功能链部署方法

针对网络功能虚拟化环境中服务功能链部署的资源开销问题, 提出一种基于流量优化的可靠服务功能链部署方法。首先, 根据虚拟链路流量变化对虚拟网络功能进行聚合, 使带宽需求大的虚拟链路部署到底层网络后, 其流量变为服务器的内部流量。其次, 综合考虑可靠性、综合时延以及拓扑属性, 利用离差最大化的多指标决策算法对物理节点进行评价, 并使用链路约束降低流量的乒乓效应。最后, 采用k-最短路径算法将虚拟链路部署到满足带宽需求的物理链路上。仿真实验表明, 所提方法在保证服务功能链可靠性的基础上, 对长期收益开销比、映射成功率、端到端时延以及带宽开销有较大优化。

Reliable service function chain deployment method based on traffic optimization

Aiming at the problem of resource cost of service function chain deployment in network function virtualization environment, a reliable service function chain deployment method based on traffic optimization is proposed. Firstly, the virtual network function is aggregated according to the traffic change of virtual link. After the virtual link with high bandwidth demand is deployed to the underlying network, its traffic becomes the internal traffic of the server. Secondly, considering the reliability, delay and topology attributes comprehensively, the multi-index decision algorithm with maximum deviation is used to evaluate the physical nodes, and the distance constraint is used to reduce the ping-pong effect of traffic. Finally, k-shortest path algorithm is used to deploy the virtual link to the physical link that meets the bandwidth demand. Simulation experiments show that this method can optimize the long-term revenue overhead ratio, mapping success rate, end-to-end delay and bandwidth overhead on the basis of ensuring the reliability of service function chain.