JaGW: A hybrid meta-heuristic algorithm for IoT workflow placement in fog computing environment

Abstract

In recent years, applications of the Internet of Things (IoT) have experienced rapid growth, driven by the widespread adoption of IoT devices in various sectors. However, these devices are typically resource-constrained in terms of computing power and storage capacity. As a result, they often offload the generated data and tasks to nearby edge devices or fog computing layers for further processing and execution. The fog computing layer is located in close vicinity of the IoT devices and comprises a set of heterogeneous fog computing nodes to supplement the capacities of resource-constrained IoT devices. The fog computing nodes often pose computational challenges for various computation-intensive tasks such as image processing application, comprises various machine learning and artificial intelligence enabled tasks. In such a scenario, finding the effective task placement for dynamic and heterogeneous applications is computationally hard. In this work, we formulate the IoT application workflow placement problem as a multi-objective optimization problem formulated as Integer Linear Programming (ILP) model with the objective of minimizing the makespan, cost of execution, and energy consumption. A hybrid metaheuristic approach is proposed that combines the strengths of the Jaya algorithm (JA) and Grey Wolf Optimization (GWO) named as JaGW to derive a sub-optimal solution. The proposed JaGW is compared with conventional GWO and other state of the art algorithms such as Ant Colony Optimization (ACO) and Particle Swarm Optimization (PSO) using the Montage scientific workflow dataset. The simulation results demonstrate that the proposed algorithm achieves an average reduction in energy consumption of 24.84% compared to JAYA, 14.67% compared to ACO, 14.65% compared to PSO, and 8.78% compared to GWO, thereby exemplifying its superior performance over other metaheuristic algorithms.