A Hybrid Filter for Accurate RSSI-Based LoRa Distance Estimation Systems in IoT Applications

The correlation between distance and Received Signal Strength Indication (RSSI) values plays a key role in Wireless Sensor Networks (WSN) positioning and localization techniques. However, when the distance between nodes is large or an obstacle creates a line of sight obstruction, stan-dard signal propagation models produce inaccurate distance estimates. Therefore, it is crucial to use filtering and signal processing techniques on the gathered raw data to reduce the noise and adapt to the network's dynamics, thereby alleviating any adverse effects. This paper develops an RSSI-based distance estimation scheme communicating with long-range LoRa radios. Therefore, a hybrid filter technique combining the Kalman filter and Simple Moving Average (SMA) filter is implemented in an outdoor environment. Results show that the error of distance estimation applying the developed technique reduces the error of distance measurement to 77 %, and 65 % compared to results with Kalman filter only and SMA filter only, respectively.