Energy-Aware Clustering in the Internet of Things by Using the Genetic Algorithm

Abstract

Internet of things (IoT) uses a lot of key technologies to collect different types of data around the world to make an intelligent and integrated whole. This concept can be as simple as a connection between a smartphone and a smart TV, or can be complex communications between the urban infrastructure and traffic monitoring systems. One of the most challenging issues in the IoT environment is how to make it scalable and energy-efficient with regard to its growing dimensions. Object clustering is a mechanism that increases scalability and provides energy efficiency by minimizing communication energy consumption. Since IoT is a large scale dynamic environment, clustering of its objects is a NP-Complete problem. This paper formulates energy-aware clustering of things as an optimization problem targeting an optimum point in which, the total consumed energy and communication cost are minimal. Then. it employs the Genetic Algorithm (GA) to solve this optimization problem by extracting the optimal number of clusters as well as the members of each cluster. In this paper, a multi objective GA for clustering that has not premature convergence problem is used. In addition, for fast GA execution multiple implementation, considerations has been measured. Moreover, the consumed energy for received and sent data, node to node and node to BS distance have been considered as effective parameters in energy consumption formulation. Numerical simulation results show the efficiency of this method in terms of the consumed energy, network lifetime, the number of dead nodes and load balancing.