Power-aware gateway connectivity in battery-powered dynamic IoT networks

The paradigm of Internet of Things (loT) is on rapid rise in today's world of communication. Every networking device is being connected to the Internet to develop specific and dedicated applications. Data from these devices, called as loT devices, is transmitted to the Internet through IoT Gateways (IGWs). IGWs support all the technologies in an IoT network. In order to reduce the cost involved with the deployment of IGWs, specialized low-cost devices called Solution Specific Gateways (SSGWs) are also employed alongside IGWs. These SSGWs are similar to IGWs except they support a subset of technologies supported by IGWs. A large number of applications are being designed which require IGWs and SSGWs to be deployed in remote areas. More often than not, gateways in such areas have to be run on battery power. Hence, power needs to be conserved in such networks for extending network life along with maintaining total connectivity. In this paper, we propose a dynamic spanning tree based algorithm for power aware connectivity called SpanloTPower-Connect which determines (near) optimal power consumption in battery-powered loT networks. SpanIoTPower-Connect computes the spanning tree in the network in a greedy manner in order to minimize the power consumption and achieve total connectivity. Additionally, we propose an algorithm to conserve power in dynamic loT networks where the connectivity demand changes with time. Our simulation results show that our algorithm performs better than Static Spanning Tree based algorithm for power-aware connectivity (Static ST) and a naive connectivity algorithm where two neighboring SSGWs are connected through every available technology. To the best of our knowledge, our work is the first attempt at achieving power-aware connectivity in battery-powered dynamic loT networks. (C) 2017 Elsevier B.V. All rights reserved.