ANLoC: An Anomaly-Aware Node Localization Algorithm for WSNs in Complex Environments.

Area of Science:

• Computer Science
• Electrical Engineering
• Network Engineering

Background:

• Accurate node location is vital for Wireless Sensor Networks (WSNs).
• Existing range-based localization methods struggle with incomplete or distorted measurements and limited noise distribution assumptions.
• Low-rank matrix recovery methods often fail with complex noise patterns beyond Gaussian or Laplacian distributions.

Purpose of the Study:

• To propose a novel Anomaly-aware Node Localization (ANLoC) method for WSNs.
• To simultaneously impute missing range measurements and detect node anomalies in complex environments.
• To address limitations of existing methods in handling wider noise distributions.

Main Methods:

• Utilizing the low-rank property of the Euclidean Distance Matrix (EDM).
• Formulating range imputation as a Robust ℓ 2,1-norm Regularized Matrix Decomposition (RRMD) model.
• Fitting complex noise with a Mixture of Gaussian (MoG) distribution and detecting anomalies using ℓ 2,1-norm regularization.
• Employing an Expectation Maximization (EM) algorithm for optimization and Multi-Dimensional Scaling (MDS) for node positioning.

Main Results:

• The proposed RRMD model effectively handles complex noise distributions.
• The ANLoC method successfully imputes missing range measurements and identifies anomalous nodes.
• Experimental results show superior performance compared to three state-of-the-art localization methods.

Conclusions:

• The ANLoC method offers a robust solution for WSN localization in challenging environments.
• The approach effectively addresses issues of missing data and node anomalies.
• ANLoC demonstrates significant improvements in localization accuracy and reliability.