An Opposition Learning-based Quantum Inspired Salp Swarm Optimization for The Multiobjective Controller Placement Problem
Sanjai Pathak, Ashish Mani, Amlan Chatterjee · 2023
This paper tackles the multiobjective controller placement problem in Software Defined Networks (SDN), a complex optimization challenge affecting network Quality of Service (QoS). In contrast to prior approaches focusing solely on propagation latency, our method considers propagation latency, inter-controller latency, and load balancing when allocating switches to controllers. We introduce EQSSA, a hybrid model that combines Elite Opposition Learning (EOL), Quantum Computing, and the Salp Swarm Optimization Algorithm (SSA). SSA is a powerful meta-heuristic for real-world optimization problems, while EOL adds cost-effectiveness and efficiency to problem-solving. Quantum-inspired techniques enhance exploration and exploitation compared to standard methods. The hybridization aims to maintain SSA's diversity and computing power. We formulate the multiobjective controller placement problem (MOCPP) based on Pareto optimality and employ EQSSA for evaluation. Experimental results, compared to standard SSA and tested on well-known benchmarks, affirm the effectiveness of our approach, demonstrating EQSSA's competitiveness in solving MOCPP and offering a promising solution to SDN controller placement challenges.