Implementation of Fuzzy Kalman Filter in Indoor Localization using Ultra-Wideband Sensor

Indoor localization is an essential technology for indoor position tracking, one of which utilizes Ultra-Wideband (UWB) sensors because of their high accuracy. However, UWB sensor readings still contain noise that reduces data reliability. To overcome this, this study proposes a solution in the form of a Kalman Filter with adaptive determination of process noise (Q) and measurement noise (R) values based on fuzzy logic. The main contribution of this study is the application of Fuzzy Kalman Filter (FKF) to dynamically adjust Q and R values, thereby increasing resistance to noise. The method used is the Mamdani fuzzy logic system integrated into the Kalman Filter, where fuzzy rules regulate Q and R updates based on variations in estimation errors. Performance analysis uses Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). The experimental results show that applying the Fuzzy Kalman Filter to UWB sensor readings can significantly reduce RMSE and MAE values. Almost the same RMSE and MAE values indicate increased accuracy and outlier reduction. In addition, using the Kalman Filter affects the estimation of tag position coordinates obtained from the Trilateration method, thereby reducing deviation and producing more stable RMSE and MAE values. This study concludes that the Fuzzy Kalman Filter successfully improves the accuracy of UWB-based indoor localization while reducing noise and outliers in measurement data.