Neurodynamic robust adaptive UWB localization algorithm with NLOS mitigation.

Area of Science:

• Robotics
• Signal Processing
• Computer Vision

Background:

• Accurate indoor localization is challenging due to mixed line-of-sight (LOS) and non-line-of-sight (NLOS) conditions.
• Existing algorithms often require prior knowledge of NLOS bias or LOS/NLOS identification, limiting their applicability.
• Robustness and accuracy in severe NLOS environments remain critical issues for indoor positioning systems.

Purpose of the Study:

• To develop a robust adaptive positioning algorithm for indoor environments with mixed LOS and NLOS.
• To address the challenges posed by the non-convexity of a proposed max-min optimization estimator.
• To improve localization accuracy and robustness, particularly in challenging NLOS scenarios.

Main Methods:

• Proposed a max-min optimization estimator and an adaptive loss function for robust estimation.
• Employed a neurodynamic approach to solve the non-convex optimization problem.
• Introduced an adaptive distance error upper bound method to enhance localization model accuracy.

Main Results:

• The neurodynamic approach effectively solved the non-convex optimization problem.
• The algorithm demonstrated superior localization accuracy and robustness compared to existing methods.
• Performance was particularly notable in severe NLOS environments, validated through simulations and real-world tests.

Conclusions:

• The proposed robust adaptive positioning algorithm offers significant improvements for indoor localization in mixed LOS/NLOS environments.
• The method's ability to operate without prior LOS/NLOS information or NLOS bias magnitude enhances its practical utility.
• This work provides a more accurate and reliable solution for indoor positioning, especially under adverse signal propagation conditions.