Initial Estimate Selection Method in Passive TDOA-Based Iterative Position Estimation Algorithms

Area of Science:

• Engineering
• Computer Science
• Signal Processing

Background:

• Iterative position estimation algorithms using Time Difference of Arrival (TDOA) are fundamental in passive localization systems like underwater acoustic and wireless sensor networks.
• The convergence and accuracy of these algorithms are critically dependent on the initial estimate, a factor often underestimated in practical applications.

Purpose of the Study:

• To present and evaluate a case-specific approach for selecting initial estimates in passive TDOA-based iterative position estimation.
• To compare the performance of two novel methods against a standard baseline approach for initial guess selection.

Main Methods:

• Simulations were performed in Python for both 2D and 3D scenarios, utilizing two distinct sensor geometric configurations.
• A grid-based analysis across a 2 × 2 km area assessed performance under noise-free and varying levels of Gaussian noise in TDOA measurements.

Main Results:

• The proposed Method 1 demonstrated significant improvements in reliability, particularly for asymmetric sensor configurations in regions sensitive to initialization.
• Both proposed methods were evaluated against a baseline where the initial guess is at the sensor formation's center.

Conclusions:

• Effective initial estimate selection is vital for enhancing the accuracy and robustness of passive TDOA position estimation.
• The findings underscore the importance of tailored initialization strategies, especially in passive systems with limited prior information or challenging sensor geometries.