A Fault-Tolerant Load Balancing Approach Using Whale Optimization Algorithm for Cloud-Fog Based IoT Applications.

Cloud computing provides storage and computation for on-demand IoT (Internet of Things) usage. The huge requests of IoT device create network congestion and late response. The real-time and delay-sensitive requests need fast processing which can be provided by fog computing closer to network edge. A compound cloud-fog model can manage both real-time and delayed requests. The possibility of distribution of real-time requests among fog nodes arises uneven load distribution which increases with reassignment of requests of fault nodes. Hence a fault tolerant load balancing algorithm based on whale optimization algorithm is proposed that can evenly distribute the load of resources of fog node optimally among both fresh requests and requests coming from faulty nodes. The metrics as makespan and percentage of successful completion of requests are evaluated and validated with other algorithms. In addition to this an IoT hardware setup is designed and checked outcomes of the metrics.