A Cooperative Hybrid Quantum-Inspired Salp Swarm and Differential Evolution for Solving CEC 2022 Benchmark and Controller Placement Problems in Software Defined Networks

Abstract

This paper introduces a Cooperative Model of Salp Swarm Optimization (CMSSO), which combines four algorithms: Salp Swarm Algorithm (SSA), Elite Opposition Learning-based SSA (EOSSA), Elite Opposition Quantum-inspired SSA (EQSSA), and Individual Dependent Approach for Differential Evolution (IDA-DE). These algorithms collaborate to tackle single-objective numerical optimization benchmarks from CEC-2022. SSA is a robust population-based metaheuristic renowned for its efficacy in practical optimization tasks. EOL and Quantum-inspired evolutionary algorithms exhibit enhanced capabilities in navigating search spaces compared to standard evolutionary algorithms. The objective of this cooperative model is to preserve the diversity and computational prowess of SSA while leveraging the strengths of these advanced algorithms. The multiobjective controller placement problem in Software Defined Networks (SDN) involves assigning switches to controllers, impacting network Quality of Service (QoS). Previous studies often focused on propagation latency for this assignment. However, our paper addresses this problem considering propagation latency between switches and controllers, inter-controller latency, and load balancing as multiobjective optimization. The experimental results confirmed the effectiveness of the proposed approach and showed that CMSSO is competitive with the standard SSA approaches.

Keywords

• Multiobjective combinatorial optimization
• SSA
• computational intelligence
• SDN
• controller placement problem