[Paper Review] Distributed Source Localization in Wireless Underground Sensor Networks

Hongyang Chen, Wentao Ouyang+1 more|arXiv (Cornell University)|Dec 17, 2011

Indoor and Outdoor Localization Technologies14 references3 citations

TL;DR

This paper proposes a distributed source localization scheme for wireless underground sensor networks (WUSNs) using robust Chinese remainder theorem (RCRT)-based time-difference-of-arrival (TDOA) estimation to improve ranging accuracy in noisy, dispersive underground channels. By leveraging a diffusion-based distributed algorithm, the method achieves high localization accuracy with low communication cost and computational complexity, outperforming local and centralized approaches in simulation while approaching the Cramér-Rao lower bound.

ABSTRACT

Node localization plays an important role in many practical applications of wireless underground sensor networks (WUSNs), such as finding the locations of earthquake epicenters, underground explosions, and microseismic events in mines. It is more difficult to obtain the time-difference-of-arrival (TDOA) measurements in WUSNs than in terrestrial wireless sensor networks because of the unfavorable channel characteristics in the underground environment. The robust Chinese remainder theorem (RCRT) has been shown to be an effective tool for solving the phase ambiguity problem and frequency estimation problem in wireless sensor networks. In this paper, the RCRT is used to robustly estimate TDOA or range difference in WUSNs and therefore improves the ranging accuracy in such networks. After obtaining the range difference, distributed source localization algorithms based on a diffusion strategy are proposed to decrease the communication cost while satisfying the localization accuracy requirement. Simulation results confirm the validity and efficiency of the proposed methods.

Motivation & Objective

• To address the challenge of inaccurate TDOA estimation in wireless underground sensor networks (WUSNs) due to dispersive underground channels and phase ambiguity.
• To improve ranging accuracy in WUSNs by applying the robust Chinese remainder theorem (RCRT) to mitigate noise-induced errors in TDOA measurements.
• To reduce communication and computational costs in source localization by developing a distributed diffusion-based algorithm instead of centralized fusion.
• To achieve localization accuracy close to the centralized optimal solution while maintaining robustness and scalability in dynamic underground environments.

Proposed method

• Uses the robust Chinese remainder theorem (RCRT) to estimate range differences from noisy TDOA measurements, overcoming phase ambiguity and noise sensitivity inherent in the classical CRT.
• Employs a diffusion strategy where sensor nodes iteratively exchange and update local estimates with one-hop neighbors, enabling distributed, robust, and scalable source localization.
• Introduces three variants of the diffusion algorithm: Diff (con) with constant weights, Diff (wei) with covariance-weighted fusion, and Diff (opt) with optimal weight selection based on estimation reliability.
• Applies statistical signal processing to estimate source position from RCRT-based range differences, using a cluster-based architecture to manage network scalability.
• Utilizes a centralized global estimator as a performance benchmark to evaluate the accuracy of distributed algorithms.
• Employs simulation-based evaluation with metrics including RMSE, CPU time, and convergence iterations to compare performance across algorithms.

Experimental results

Research questions

• RQ1 Can RCRT-based TDOA estimation significantly improve ranging accuracy in WUSNs under high noise and multipath fading?
• RQ2 How does the performance of distributed diffusion-based localization compare to centralized and local estimation in terms of accuracy and communication cost?
• RQ3 What is the impact of network topology, number of clusters, and measurement noise on the convergence and accuracy of diffusion algorithms?
• RQ4 Can the proposed distributed algorithm achieve localization accuracy close to the Cramér-Rao lower bound while minimizing computational and communication overhead?

Key findings

• The RCRT-based TDOA estimation method effectively mitigates noise-induced errors, enabling robust range difference estimation even with bounded remainder errors.
• The diffusion-based distributed localization algorithm (Diff) achieves localization accuracy close to the centralized global estimator, with RMSE values approaching the Cramér-Rao lower bound in simulations.
• Diff (opt) and Diff (wei) outperform Diff (con) due to their use of estimation covariance matrices for adaptive weighting, with Diff (opt) showing the best accuracy but higher CPU cost.
• Diff (con) is the most communication- and computation-efficient, with CPU time nearly equal to local estimation, while Diff (wei) and Diff (opt) require slightly more time due to weight computation.
• As the number of clusters (N) increases beyond a certain point, performance degrades due to geometric dilution of precision, especially when the source lies outside the convex hull of sensor clusters.
• An optimal value of the weighting parameter γ exists for Diff (wei), and performance remains superior to Diff (con) across a wide range of γ values, demonstrating robustness to parameter selection.