Efficient Distributed Method for NLOS Cooperative Localization in WSNs.

Area of Science:

• Robotics and Automation
• Wireless Sensor Networks
• Signal Processing

Background:

• Cooperative localization accuracy in wireless sensor networks (WSNs) is significantly reduced in non-line-of-sight (NLOS) environments.
• Existing methods often struggle with computational speed, limiting their practical applicability.

Purpose of the Study:

• To propose a novel distributed cooperative localization method for WSNs that addresses the challenges posed by NLOS environments.
• To enhance both the accuracy and computational efficiency of cooperative localization algorithms.

Main Methods:

• Developed a convex model based on projection relaxation, suitable for scenarios lacking prior information on NLOS connections.
• Designed an efficient decomposed formulation for the convex counterpart.
• Implemented a parallel distributed algorithm using the alternating direction method of multipliers (ADMM) and a proximal algorithm to accelerate convergence.

Main Results:

• The proposed ADMM-based algorithm significantly improves computational speed compared to existing methods.
• Numerical simulations confirm the approach's superiority in both processing speed and accuracy within NLOS scenarios.
• The method effectively handles situations with unknown NLOS connections.

Conclusions:

• The developed distributed cooperative localization method offers a practical and efficient solution for WSNs operating in challenging NLOS environments.
• The combination of projection relaxation, ADMM, and proximal algorithms provides a robust framework for high-speed and accurate localization.