Novel Gain-Optimized Two-Step Fusion Filtering Method for Ranging-Based Localization Using Predicted Residuals.

Area of Science:

• Wireless Sensor Networks
• Localization Algorithms
• Signal Processing

Background:

• Distance-based positioning in wireless sensor networks suffers from disturbance and modeling errors.
• Inaccurate initial filtering values can lead to significant estimation errors or divergence in positioning.
• Existing algorithms struggle with achieving high accuracy and robust performance.

Purpose of the Study:

• To propose a two-stage fusion filtering positioning algorithm that addresses limitations of current methods.
• To enhance the accuracy and reliability of target localization in wireless sensor networks.
• To mitigate issues caused by disturbances, modeling errors, and initial value inaccuracies.

Main Methods:

• Constructing a pseudo-linear equation using time of arrival (TOA) and multipath delay via parameterization.
• Applying the weighted least squares (WLS) method to determine an initial target position estimate near the Cramér-Rao lower bound (CRLB).
• Employing a Kalman filtering algorithm with reconstructed Gaussian white noise statistics for precise positioning.

Main Results:

• The proposed algorithm achieves an average positioning accuracy improvement of 28.57% compared to other initial positioning algorithms.
• Demonstrates superior filtering performance, leading to more reliable target localization.
• The initial value estimation approaches the Cramér-Rao lower bound (CRLB).

Conclusions:

• The two-stage fusion filtering algorithm effectively improves positioning accuracy and filtering performance in wireless sensor networks.
• The method provides a robust solution for target localization challenges posed by disturbances and modeling errors.
• This approach offers a significant advancement for distance-based positioning in challenging wireless environments.