Socio-Physical Energy-Efficient Operation in the Internet of Multipurpose Things

Abstract

Multipurpose devices have emerged as part of modern Internet of Things (IoT) ecosystems. Such nodes are able of interchanging their operation between different sensing modes providing a large mixture of information towards various IoT applications. In this paper, a novel framework is introduced to govern and properly define the dynamic operation of a multipurpose device network deployment. Initially, the problem of socio- physical energy-efficient device sensing mode selection is confronted. Each multipurpose device acts as a learning automaton and through a machine learning mechanism selects the most appropriate operation mode, in terms of maximizing the revenue/cost relation of the provider. In addition, towards improving the communication efficiency, a coalition formation mechanism among the nodes is proposed, which considers: (a) nodes' spatial proximity reflecting physical conditions such as channel quality, (b) energy availability, and (c) operation mode correlation between multipurpose devices expressing social metrics. Given the mode selection and coalition formation among nodes, a distributed utility-based power control mechanism is proposed to determine each device's optimal transmission power in a Non- Orthogonal Multiple Access (NOMA) wireless network environment in order to fulfill its Quality of Service (QoS) prerequisites. The performance of the proposed approach is evaluated through modeling and simulation under several scenarios, and its superiority is demonstrated.