Geolocation techniques for LoRa technology: a systematic mapping

Abstract

Native LoRa geolocation is a commercially emerging feature, inherent to the LoRa wireless technology, which however has important limitations. The main limitation is low accuracy: 20-200m for gateways equipped with high-resolution clocks, and 1-2km for simpler gateways. In the last years, researches have been carried out with the main aim of improving these accuracy levels. Thus, this systematic mapping obtains an overview of the state-of-the-art of native LoRa geolocation techniques. A total of 43 papers are mapped, published between 2016 and 2021. Multilateration TDoA/RSSI techniques are the most used, along with a wide range of algorithms that calculate the geographical coordinates, such as analytical methods, statistical methods, machine learning, and fingerprinting. Only 23% of the papers are from real LoRaWAN networks, highlighting a research gap in this regard. Overall, improvements in accuracy levels are observed in virtually all mapped works: 25m on average for simulated ones or experimented in controlled environments, and 300 m for tested in real signal propagation environments, such as large urban areas.