LoRaWAN Cattle Tracking Prototype With AI-based Coverage Prediction

Abstract

This work describes a low-cost tracking prototype tested in São Paulo, Brazil. Its use can prevent cattle theft, determine the distance traveled by the animal, estimate the amount of feed available daily, transmit information about the animal's health, such as temperature, and anticipate veterinary care. Such technology can also be used in the control system of hydraulic pumps, and monitoring water fountain levels, forming a remote management ecosystem. Using the collected data by the system, an artificial intelligence algorithm, long short-term memory (LSTM) network, has been implemented to predict the received signal strength indication (RSSI). In this way, we show that the prototype is also helpful for planning mobile communications networks. For example, it is possible to determine the coverage of the LoRA signal from a few measurements, especially in free-space scenarios with line-of-sight.