Cluster Head Selection Using Multi-Dilation Convolutional Neural Network Optimized with BCMO for IoT Networks

Abstract

Generally, more Internet of Things (IoT) based sensor networks are used for transportation and traffic signals in a smart city. The need of IoT based networks are increasing day by day. The main issue of IoT networks is poor energy optimization. Offering energy optimization in the IoT will present a number of difficulties. The researchers try to extend the network lifetime by reducing energy consumption. Clustering is a good solution to reduce the consumption of energy in a network. In IoT along wireless sensor networks (WSNs), a cluster is a group of nodes. Using these clusters, Cluster Head (CH) is formed. CH collects information from other nodes and is connected to the Base Station (BS). In this paper, Cluster head selection using a Multi-Dilation Convolutional Neural Network optimized with Balancing Composite Motion Optimization (BCMO) is proposed. The major goal of this work is to increase the normalized energy and optimize the weight, temperature, delay, distance between all nodes, and weight. In this paper, the Multi-Dilation Convolutional Neural Network optimized with Balancing Composite Motion optimization is proposed for energy optimization (CHS-MDCNN-BCMO-IoTN). The efficiency of the proposed CHS-MDCNN-BCMO-IoTN approach is compared with other existing methods in terms of normalized energy, load, temperature, and number of cluster head formations. This CHS-MDCNN-BCMO-IoTN method attains 31.45%, 30.98% and 27.24% highest number of alive nodes, 26.16%, 27.23% and 33.62% lower number of dead nodes when compared with existing methods.