SDN Path Recovery Scheme Using Bionic-Based Self-Healing Mechanism
Changlong Yan, Nan Sun, Tao Li, Zhongyuan Qin, Aiqun Hu · 2023
Software Defined Networking (SDN) is a novel and highly adaptable network framework. In case there is a link failure, the controller will recalculate the appropriate forwarding path and distribute new rules to the switches on the path. It will result in long recovery time and degraded efficiency in link transmission. In this paper, we propose a bionic-based self-healing method for SDN path recovery. In our scheme, the human genetic mechanism is imitated by calculating the primary/backup paths using an improved genetic algorithm. We also design a set of flow table rules referring to the human blood vessels by fast failover group table. MPLS is pre-binded for each path, therefore when a link failure occurs, the switch changes the MPLS of the packet header and performs limited backtracking to reroute the packet. Experimental results show that our scheme can effectively reduce the switch flow table overhead and improve the self-healing capability of data links.