The Power of Models: Modeling Power Consumption for IoT Devices

TLDR

Low‑energy IoT technologies still lack the reliability required for industrial wireless deployments, making efficient energy dimensioning difficult amid the uncertainty of energy harvesting. The study aims to model and dimension IoT device energy consumption before deployment to balance cost, lifetime, and available energy. The authors propose a system‑level power consumption model that aggregates communications, sensing, and processing costs and relies only on empirically measurable platform and application parameters. The resulting framework enables early analysis of energy life‑cycles, quantifies the impact of application parameters, and clarifies tolerance margins and trade‑offs.

Abstract

Low-energy technologies in the Internet of Things (IoTs) era are still unable to provide the reliability needed by the industrial world, particularly in terms of the wireless operation that pervasive deployments demand. While the industrial wireless performance has achieved an acceptable degree in communications, it is no easy task to determine an efficient energy-dimensioning of the device in order to meet the application requirements. This is especially true in the face of the uncertainty inherent in energy harvesting. Thus, it is of utmost importance to model and dimension the energy consumption of the IoT applications at the pre-deployment or pre-production stages, especially when considering critical factors, such as reduced cost, life-time, and available energy. This paper presents a comprehensive model for the power consumption of wireless sensor nodes. The model takes a system-level perspective to account for all energy expenditures: communications, acquisition and processing. Furthermore, it is based only on parameters that can empirically be quantified once the platform (i.e., technology) and the application (i.e., operating conditions) are defined. This results in a new framework for studying and analyzing the energy life-cycles in applications, and it is suitable for determining in advance the specific weight of application parameters, as well as for understanding the tolerance margins and tradeoffs in the system.