An efficient and autonomous scheme for solving IoT service placement problem using the improved Archimedes optimization algorithm

The ever-increasing growth of the number of Internet of Things (IoT) devices connected to the network has led to the emergence of cloud computing shortcomings such as delay, storage and bandwidth. Fog computing has been developed as an emerging computing paradigm to overcome the challenges of cloud computing. This paradigm can support delay-critical and computationally intensive applications by pro-viding resources at the network edge. As fog nodes appear with limited resources, IoT service placement schemes can improve the Quality of Service (QoS) and system performance. In general, the mapping between fog nodes and IoT services is known as the Service Placement Problem (SPP) in fog computing. To date, various meta-heuristic approaches have been introduced to solve SPP, but few are known in the research society due to computational complexity. Hence, this study proposes an efficient and autono-mous scheme to solve SPP using meta-heuristic approaches with shared parallel architecture that can overcome the problem complexity. Specifically, we use the Archimedes Optimization Algorithm (AOA) as a new meta-heuristic approach inspired by the physics law of Archimedes' Principle. The proposed scheme, as SPP-AOA, formulates SPP as a multi-objective problem and performs the placement of auton-omous services on distributed fog domains. Our main concerns in SPP are related to resource utilization, service cost, energy consumption, delay cost and throughput. SPP-AOA performs placement based on extracting resource distribution over time, which can save more resources to handle future requests. The effectiveness of the proposed scheme has been proven through evaluation on Barabasi-Albert net-work topology. Compared to the state-of-the-art methods, SPP-AOA deploys an average of 5% more ser-vice in fog, lowers costs by 2% and waiting time by 20%, and reduces placement on the cloud by 14%. (c) 2023 Published by Elsevier B.V. on behalf of King Saud University.