Abstract
The wide adoption of edge computing has introduced several issues such as load balancing, resource provisioning, and workload placement as optimization problems. Particle swarm optimization (PSO) is a nature-inspired stochastic optimization algorithm, whose objective is to iteratively improve the solution of a problem over a given objective. The distribution of PSO to the edge would result in the transfer of resource-intensive computational tasks from the cloud to the edge, leading to more efficient use of existing resources. However, it introduces challenges related to performance and fault tolerance, due to the resource-constrained edge environment with a high probability of faults. We introduce multiple distributed synchronous variants of the PSO algorithm built on top of the Apache Spark distributed computing framework and Kubernetes container orchestration platform. These variants of the algorithm aim at addressing the performance and fault tolerance problems introduced by the execution in an edge network. A PSO algorithm that distributes the load across multiple executor nodes can effectively realize coarse-grained parallelism, thus can obtain a significant increase in performance, but also more fault tolerance and scalability.