Soler-Fernández, Juan LuisRomera, OmarDiéguez Barrientos, ÀngelPrades García, Juan DanielAlonso Casanovas, Oscar2026-01-162026-01-162026-01-021424-8220https://hdl.handle.net/2445/225675Energy efficiency is a key requirement for Internet of Things (IoT) nodes, particularly in applications powered by energy harvesting that operate without batteries. In this work, we present a parametric power model of a LoRa transceiver (Semtech SX1276) aimed at ultra-low power remote sensing scenarios. The transceiver was characterized in all relevant states (startup, transmission, reception, and sleep), and the results were used to build a state-based model that predicts average power consumption as a function of transmission power, sleep strategy, packetization, and input data rate. Experimental validation confirmed that the cubic fit for transmission peaks achieves a determination coefficient of 0.99, while reception is added as a constant consumption. The model was implemented in a Python simulator that provides mean, best-case, and worst-case estimates of system power consumption, and it was validated in an ASIC-based sensor node demonstration, with predictions within 10% of measured values. The framework highlights the trade-offs between energy efficiency and robustness (e.g., minimal SF and no CRC vs. higher spreading factors and error-control) and supports the design of custom controllers for ultra-low power IoT nodes as well as more energy-permissive applications.17 p.application/pdfengcc-by (c) Soler-Fernández, J.L. et al., 2026http://creativecommons.org/licenses/by/4.0/Python (Llenguatge de programació)InternetRecol·lecció d'energiaPython (Computer program language)InternetEnergy harvestinginfo:eu-repo/semantics/article7635732026-01-16info:eu-repo/semantics/openAccess