RWP-NSGA II: Reinforcement Weighted Probabilistic NSGA II for Workload Allocation in Fog and Internet of Things Environment

The explosion of the IoT and the immense increase in the number of devices around the world, as well as the desire to meet the quality of service in the best way possible, have challenged cloud computing. Fog computing has been introduced to reduce the distance between the IoT and the cloud and to process time-sensitive tasks in an efficient and speedy manner. IoT devices can process a portion of the workload locally and offload the rest to the fog layer. This workload is then allocated to the fog nodes. The distribution of workload between IoT devices and fog nodes should account for the constrained energy resources of the IoT device, while still prioritizing the primary objective of fog computing, which is to minimize delay. This study investigates workload allocation in the IoT node and the fog nodes by optimizing delay and energy consumption. This paper proposes an improved version of NSGA II, namely, reinforcement weighted probabilistic NSGA II, which uses weighted probabilistic mutation. This algorithm replaces random mutation with probabilistic mutation to enhance exploration of the solution space. This method uses domain-specific knowledge to improve convergence and solution quality, resulting in reduced delay and better energy efficiency compared to traditional NSGA II and other evolutionary algorithms. The results demonstrate that the proposed algorithm reduces delay by nearly 2 s while also achieving an improvement in energy efficiency, surpassing the state of the art by nearly 3 units.