Abstract
With the edge computing paradigm bringing processing power and storage capabilities closer to the edge of the network, the need for optimal workload placement strategies for distributed applications arises. This workload placement must cope with the increasing complexity of the Internet of Things (IoT) sensor-actuator infrastructure and be able to satisfy the requirements of each application in a volatile and heterogeneous edge infrastructure. Particle Swarm Optimization (PSO) is a well-established non-deterministic evolutionary algorithm, based on bird behavior; whose objective is to iteratively improve the solution of a problem over a given objective. The algorithm in PSO searches the solution space by creating a population of solution candidates known as particles and moving them with simple mathematical functions. In this study, we introduce a Binary Multi-Objective Particle Swarm Optimization (BMOPSO) algorithm to address the workload placement problem in Edge Computing. By combining these modifications to the heuristic, we can efficiently make workload placement decisions based on specific predefined objectives. The experimental result of comparative simulations shows the benefits of the proposed BMOPSO algorithm versus a baseline, testing variations in the number of applications to be allocated.