Computational Intelligence based Optimal Service Placement in Fog-enabled IoT Networks

One of the crucial tasks in fog-enabled Internet of Things (IoT) networks is the strategic allocation and deployment of computational services. Service placement plays a powerful role in improving system performance, efficiency, and resource utilization. It also affects cost, Quality of Service (QoS), network usage, and customer satisfaction. Several studies have aimed to strengthen service allocation in the past, but this is a challenging problem for the computing research community. We observed that meta-heuristic-based computational intelligence approaches in this domain show effective performance. In this study, we propose a novel computational intelligence approach that is designed based on the hybridization of Levy and the genetic algorithm for optimal service placement in fog-enabled IoT networks. Hybridization aims to leverage the strengths of both algorithms to overcome their disadvantages and achieve the best performance in energy efficiency and QoS. Levy has adopted its solution using the randomized solution strategy, and the genetic algorithm will be explored to help in a better search of the solution space. In this way, we are going to combine the advantages of both algorithms to escape from local optima and eliminate disadvantages such as shallow exploration of the entire space in the case of using traditional optimization algorithms. We conducted experiments using iFogSim-based simulated fog computing environments, which revealed a promising outcome in terms of QoS. In terms of service placement, our proposed computational intelligence-based approach outperformed baselines across all evaluated aspects, demonstrating superior overall performance. This indicates that the proposed computational intelligence approach is beneficial to offer optimal service placement in fog-enabled IoT networks.