Energy Efficient Data Routing for IoT Networks Using Genetic Algorithm

Internet of Things (IoT) applications have gained wide attention due to their potential applications in healthcare, industry, and the home. These systems are aided by low-cost battery-powered sensor nodes that collect valuable information which is then used for autonomous decision making. However, how best to gather the data at the base station remains a challenge. In this paper, we present a Genetic Algorithm (GA) based routing protocol that determines near optimal routes for battery-powered nodes to forward data to a remote base station. The algorithm is run by the base station in a centralized manner where nodes only have their nearest one hop neighbor information. We develop a multi-objective fitness function that focuses on balancing data load on a router by considering the potential incoming data and number of connections of a router while also focusing on the shortest path and energy efficiency. Simulation results show that the proposed approach can improve the battery life and data loss of an IoT network by more than 6% and 9% respectively, compared to existing approaches that consider multipath routing. In addition, by leveraging routing information, our proposed technique successfully identifies the nodes with higher energy consumption rate and reduces energy consumption by 12% compared to other multipath routing protocols.