JIITA, vol. 2, no. 4, pp.197-202, Dec, 2018
DOI: soon
Performance Analysis of the Ranging-based Localization Methods and CKF-based Performance Enhancement
Seong Yun Cho *
School of Robotics Engineering, Kyungil University, Gyeongsan, Korea
Abstract: In this paper, the performance of the wireless localization methods using the ranging measurements is analyzed. The ranging measurements in indoor environments may include non-Gaussian errors caused by the NLOS (Non-Line-of-Sight) error or calibration error. The ranging errors may be always positive because it is caused by the additional propagation path or the additional circuit path for signals. The localization performance depends on the measurement accuracy and localization methods. When the ranging measurements contain the non-Gaussian errors, in this paper, the localization performance is analyzed according to the localization methods including the model-free methods such as ILS (Iterative Least Squares), DS (direct Solution), and DSRM (Difference of Squared Ranging Measurements) methods, and model-based Kalman filtering such as CKF (Cubature Kalman Filter). The analysis results show that the DSRM method has better performance than the other model-free methods when the ranging measurements have positive non-Gaussian errors. Also, it shows that the CKF-based filtering has better performance than the model-free methods.
Keyword: model-free localization; model-based Kalman filtering; DSRM; CKF
Full Paper: