Effective Communication between IoT Devices using Machine Learning

Abstract

The Internet of Things (IoT) consisting of several interconnected sensors capable of transferring the data over the network without human involvement. All these interconnected sensors can be managed through internet remotely. The Sensors connected in the networks are deployed randomly and are mobile in nature. The loss of sensor’s energy occurs due to the mobility nature, communication between the sensors and operations performed. The perfectly suitable sensor nodes as relay nodes and Cluster Heads (CH), and energy efficient path to the Base Station (BS) should be selected at each and every stage in the IoT network for effective communication. This paper presenting the method for effective communication between IoT Devices using ML comprising of Sensor Deployment, Clustering, Data Aggregation, Data Routing, and Reconstruction. In this paper for maximum coverage across the BS, relay sensors are deployed in hexagonal fashion. The sensors in the network transmit the diverse data and data aggregation is performed at CH level by machine learning algorithm that is Particle Swarm Optimization (PSO). The Ant Colony Optimization (ACO) algorithm is used to effectively routing the data over the network and dimensionality of the data is reduced with the Principle Component Analysis (PCA). The effective communication is established here by reducing the residual nodes. The total of 600 mobile sensor nodes and 12 relay sensor nodes are considered in the simulation of this work by the Network Simulation-3 (NS-3), only 42 joules energy is consuming over 200 milliseconds with network life time of 3000 milliseconds.