Toward low-delay and low-cost controller placement in SDNs: a hybrid heuristic method
Siyuan Peng, Dongshuai Niu, Kezhong Jin, Zhenzhou Tang · 2023
The controller placement problem is an significant issue in software defined networking and has a critical impact on the performance of the network. So, this paper investigates the optimal controller placement (OCP) problem with the objectives of minimizing the controller-controller delay, controller-switch delay and deployment cost on the premise of satisfying the full coverage of the switch, maximum allowable delay and the maximum allowable load of the controller. This is an NP-hard optimization problem. To solve it, a hybrid heuristic algorithm named SCSMA that combines the slime mould algorithm and sine cosine algorithm is proposed in this paper. We comprehensively compare the performance of SCSMA with six state-of-the-art algorithms on 10 benchmark functions. The results comprehensively illustrate the excellent global optimization capability of the algorithm. After that, this paper uses weighted summation to merge the three objectives into one, and develops an OCP method based on SCSMA. The experimental simulation results demonstrate that the proposed SCSMA-OCP method is able to provide very effective solutions to OCP problems.