Workload Allocation in Fog Environment Using Multi-Objective Evolutionary Algorithms for Internet of Things

The continuous rise in the number of IoT devices has led to an increasing importance of the fog computing paradigm. Part of the workload that should be processed is executed locally on the IoT device and the rest is offloaded and allocated to the fog nodes. This workload allocation decision should be done in a way that provides the lowest delay but while taking into account the energy consumption as well. This study presents an optimization of the workload allocation that minimizes delay and power consumption using the multi-objective evolutionary algorithms, namely, NSGA II, R-NSGA II, NSGA III, R-NSGA III and CTAEA. The experiments involve two scenarios, full transmission power of the IoT device, and half of its transmission power with varying workload sizes. The results manifested the superior performance of NSGA III and CTAEA in optimizing the allocation of tasks in fog computing environments. By demonstrating NSGA III and CTAEA’s effectiveness, this findings not only advance the understanding of evolutionary algorithms but also provide practical insights for optimizing fog computing systems. This research has broader implications for improving the efficiency and performance of fog computing applications, with potential applications across various scenarios in the field.