Multi-Objectives Firefly Algorithm for Task Offloading in the Edge-Fog-Cloud Computing

Abstract

This study proposes a two-stage strategy to ensure the successful execution of critical tasks. In the first stage, an Enhanced Task Offloading (ETO) algorithm is introduced to determine the appropriate computing layer—edge, fog, or cloud—for offloading incomplete tasks. The algorithm makes this decision by assessing the availability of idle computing resources relative to the task’s computational requirements. Additionally, it verifies the status of the server (on/off) before offloading; if the server is unavailable, the algorithm proceeds to check the next layer. In the second stage, the strategy employs a Multi-objective Firefly (MFA) algorithm to assign the optimal computational device within the selected layer. Experimental simulations compare the proposed strategy with a benchmark task offloading algorithm. The results demonstrate the superiority of the proposed strategy, achieving reductions in energy consumption and delay and maximizing resource utilization compared to the baseline algorithms.

Keywords

• Task offloading
• optimization
• firefly algorithm
• metaheuristic
• edge-fog-cloud computing
• Internet of Things