Adaptive Evolution Algorithm Based on Hypervolume Awareness for Controller Placement

Abstract

Multi-controller SDN realizes logically centralized control by deploying multiple controllers in the network. A reasonable number of controllers and suitable deployment locations are beneficial for optimizing the performance of the entire network. Compared with the single-objective model, the multi-objective optimization model can obtain a set of controller deployment solutions through the multi-objective optimization algorithm, thus providing more comprehensive solutions. In this paper, an actual many-objective model is built to optimize controller deployment by considering the propagation delay between controllers and switching nodes, propagation delay between controllers, controller load difference, reliability of the control network, and deployment cost. To solve this model, we propose an algorithm by specially designing the hybrid initialization method to generate an initial population that balances diversity and convergence. After that, we design the recall mechanism based on hypervolume awareness, the deduplication elite archive storage mechanism, the adaptive evolution mechanism, and the opposition-based learning strategy. These mechanisms are particularly constructed for the proposed algorithm to solve the problems in the evolution process and to improve the global search ability of the algorithm for obtaining superior non-dominated solution sets. Finally, we validate the effectiveness and generality of the proposed algorithm by comparing its non-dominated solution with those of other algorithms in the Cogentco network from various aspects.