Heuristic Approach to Indoor Localization Using LoRa RSSI Measurements

Abstract

This research presents a heuristic approach for indoor localization using standard LoRa modules operating at 915MHz. To overcome the challenges presented by signal attenuation, multipath propagation, and environmental variability. The proposed method combines RSSI based distance estimation with a path loss exponent tuned empirically for different environments. A trilateration algorithm based on OLS is employed to estimate target positions, and performance is enhanced using filtering techniques such as Median Filter(MF) and Moving Average Filter (MAF). Additionally, two receiver geometries were analyzed to assess the robustness of the proposed method under different geometric configurations. To complement the OLS estimator, a Weighted Least Squares (WLS) method was also implemented usinga Gauss–Newton optimization approach. While WLS shows promising results, further refinement of the covariance matrix Q is identified as a direction for future work. These findings underscore the potential of the approach as a low cost, scalable solution for precise indoor localization in complex environments. Experimental evaluations conducted in various laboratory environments demonstrated that the optimized parameters yield a substantial reduction in positioning error. Performance was quantified using Mean Square Error (MSE), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE) metrics, with MSE values as low as 0.2491 m in some settings, across all scenarios without filters, and achieving 0.07 m, with appropriate window size for MF and MAF. A brief review of results shows that an MF and MAF with window size Wn = 7 provides consistently adequate accuracy while keeping computational costs low.