Energy Consumption Evaluation of LPWAN: A Stochastic Modeling Approach for IoT Systems

Abstract

Energy consumption has been a well-known concern on computing systems, and due to the emergence of IoT applications, particular attention has been devoted to network technologies. The energy for a communication transceiver component to send one single bit can be 1500 to 2000 times higher than the energy required to execute a single instruction, and many IoT applications adopt energy-constrained networks, such as low-power wide-area networks (LPWANs). Thus, mechanisms for assessing energy consumption of the network physical layer for IoT applications are prominent for system design. Despite the importance of assessing the communication transceiver’s energy consumption, to the best of our knowledge, just a few works propose models for evaluating LPWAN and WSN transceivers independently of technology with probabilistic packet loss behavior. This paper presents a modeling approach based on Stochastic Petri nets (SPN) formalism family capable to evaluate the energy consumption for LPWAN and WSN networks, focusing on the communication transceivers and channel with packet loss probabilistic behavior. The proposed models have been validated using Lora physical layer chip with an error of 1.6% to transmit data and 1.3% to receive data. The experimental results show the feasibility of the proposed models to estimate a node and network energy consumption.