Butterfly-PSO Based Energy Efficient Data Analysis of Wireless IoT Devices

Abstract

Paper collecting data from various sources for research observation, security, etc. are depend on IOT networks. As IOT device are remotely which transform information from nearby area and lifespan of this network rely on energy uses for communication. So this paper proposed a neural network and genetic algorithm combination for increasing the life span of the network. Error Back Propagation neural network was trained to identify best set of nodes for the cluster center selection. This machine learning based data selection increase the cluster selection accuracy of the BFPSO (Butterfly Particle Swarm Optimization). As combination get reduce by neural network data analysis so less number of population need to be developed for BFPSO algorithm which ultimately increase the accuracy of device selection. Various set of region size and number of nodes were developed to evaluate proposed model. Comparison of proposed model NN-BFPSO-CHS (Neural Network Butterfly Particle Swarm Optimization based Cluster Head Selection) was done with previous existing methods on different evaluation parameters and it was obtained that proposed model has improved all set of parameters.