Fast Recovery Method for SDN Faulty Links Based on Adaptive Genetic Algorithm

Aiming at the strict low-latency data transmission requirements after fault link recovery in the smart grid wide-area measurement system, this paper proposes a fast restoration method for faulty links based on adaptive genetic algorithm (FRMFL_AGA) in the software-defined network environment. FRMFL_AGA optimizes the two levels of backup path construction and backup path installation. Adaptive genetic algorithm is used to calculate the shortest backup path, and the automatically adjusted crossover probability and mutation probability are used in the genetic algorithm training process to reduce the data transmission delay during the restoration of the faulty link. For the backup path output after algorithm training converges, use the backup path installation method to complete the distribution of flow entries, effectively reducing the storage resource consumption of OpenFlow switches. Experimental results show that, compared with other faulty link recovery methods, FRMFL_AGA reduces the average fault recovery delay by about 13.27%, reduces the number of forwarding rules generated by about 17.69%, and increases the success rate of faulty link recovery by about 12.42%.