EFog-IoT: Harnessing Power Consumption in Fog-Assisted of Things

The ever-increasing use of Internet of Things (IoT) devices like smartphones, PDAs, smartwatches, etc. by the users has also drastically increased the volume of data that needs to be processed by the cloud servers. The cloud servers are very powerful and are capable of processing data at once. However, being a centralized paradigm and the existence of the physical gap from the IoT layer, it is incapable of bulk processing of data thereby resulting in latency overhead, increased power consumption, and increased service rate. This work proposes an energy-efficient method with the introduction of Fog computing as an intermediate layer between the IoT and the Cloud for computing the tremendous data generated by the IoT devices in a distributed manner in order to reduce the power consumption. Here, a Multi-level Feedback Queue is used for target node classification for minimizing the service rate, and the Fuzzy C-means++ approach is applied for clustering of available fog nodes for parallel processing. This research proposes a parallel algorithm devised for task scheduling in the fog nodes which results in power consumption while maximizing the resource utilization. Simulation results show that the proposed method performs at 98% efficiency in processing all the tasks generated by the users.