A Mechanism for Analyzing the Energy Consumption and Environmental Impact of LoRa Sensors Over Time
TL;DR
This transient analysis demonstrates that sustainability assessment in LoRa networks must consider the interactions among active consumption, reception overhead, residual idle drain, and information density throughout the device's operational cycle, providing a reliable basis for early scenario evaluation and energy‐aware planning for large‐scale IoT deployments.
Abstract
Long Range (LoRa) sensors are widely adopted in Internet of Things (IoT) applications due to their combination of long‐range communication and low energy consumption. However, the continuous, large‐scale operation of these devices raises concerns about cumulative energy consumption and environmental impacts over time. Evaluations based solely on prototypes and field tests tend to be costly, time‐consuming, and poorly reproducible, hindering the systematic comparison of multiple operational scenarios. To address these challenges, this work proposes a mechanism for analyzing the energy consumption and environmental impact of LoRa sensors over time, focusing exclusively on the sensor node. The goal is to support sustainable planning of large‐scale LoRa deployments by enabling the temporal evaluation of battery discharge, transmitted packets, and CO emissions under different message generation rates and physical transmission parameters. The mechanism is developed using Stochastic Petri Nets (SPN), which model the alternation between the operating states of a LoRa sensor. Environmental impact is quantified by converting accumulated energy consumption into CO emissions using a global average carbon emission factor. The SPN model is cross‐validated against a discrete‐event simulation implemented in NS‐3. The cross‐validation indicated close agreement between the SPN model and NS‐3, with mean absolute percentage errors below 0.3% across all evaluated metrics. The packet generation rate is shown to govern both sensor lifetime and the temporal distribution of CO emissions, with the per‐packet carbon footprint increasing under low‐traffic scenarios due to the higher share of idle consumption. The Spreading Factor (SF) exhibits a monotonic effect on energy consumption, per‐packet CO, and battery drain, with SF12 emitting approximately 8.55 times more CO per packet than SF7. When projected to networks of thousands of devices, this behavior produces a multiplicative effect on the collective carbon footprint. Ultimately, a multidimensional analysis via a radar chart illustrates the trade‐offs among lifetime, energy efficiency, data throughput, and carbon footprint across the evaluated scenarios. The proposed SPN‐based model effectively characterizes the temporal evolution of energy consumption and environmental impact of LoRa sensors, revealing cumulative dynamics that average metrics or isolated final values would not capture. This transient analysis demonstrates that sustainability assessment in LoRa networks must consider the interactions among active consumption, reception overhead, residual idle drain, and information density throughout the device's operational cycle, providing a reliable basis for early scenario evaluation and energy‐aware planning for large‐scale IoT deployments.