Hierarchical SDN Multi-controller Placement Strategy Based on Improved Aquila Optimizer

The improved Aquila optimizer is used to solve the problems of large network latency and imbalance controller load caused by placing multiple controllers in hierarchical architecture software-defined networking. Random opposition-based learning was added to the initial population, and dynamic reverse learning was added to the high-altitude flight phase of global exploration to improve the global exploration capability of the Aquila optimizer. Toward shorten the feasibility of the optimal individual trap the local optimal, the Gaussian walk strategy is added to the local development to adjust the convergence speed of the Aquila optimizer. Through the experimental verification of China Telecom network topology, the strategy solved by the improved Aquila optimizer can maintain the premise of the minimum network total latency, realize the load balance of the network, and the performance index is better than the random algorithm, K-means algorithm, particle swarm algorithm.