A Fusion Localization Method based on a Robust Extended Kalman Filter and Track-Quality for Wireless Sensor Networks

Area of Science:

• Wireless Sensor Networks (WSNs)
• Localization Techniques
• Signal Processing

Background:

• Wireless sensor networks (WSNs) are essential technologies integrating sensing, embedded computing, and communication.
• Localization is a key technique for WSNs, significantly impacting their application prospects.
• Non-line-of-sight (NLOS) propagation in complex environments like indoors causes positioning errors, degrading traditional Extended Kalman Filter (EKF) performance.

Purpose of the Study:

• To propose a novel localization method to mitigate NLOS errors in WSNs.
• To enhance positioning accuracy in both line-of-sight (LOS) and NLOS conditions.
• To develop a robust fusion algorithm for improved WSN localization.

Main Methods:

• A robust extended Kalman filter and track-quality-based (REKF-TQ) fusion algorithm is proposed.
• EKF and REKF are used in parallel to obtain initial location estimates.
• Kalman filters (KFs) with new observation vectors and equations further filter estimates, which are then fused using a track-quality-based algorithm.

Main Results:

• The REKF-TQ algorithm demonstrates superior positioning accuracy compared to EKF and REKF in NLOS environments.
• The proposed algorithm achieves higher accuracy than the interacting multiple model (IMM) algorithm using EKF and REKF.
• The fusion approach effectively mitigates the impact of NLOS errors on WSN localization.

Conclusions:

• The REKF-TQ algorithm offers a significant improvement in WSN localization accuracy, particularly in challenging NLOS conditions.
• This method provides a robust solution for achieving high-performance positioning in diverse WSN environments.
• The track-quality-based fusion strategy is effective in combining estimates for enhanced localization precision.