A Robust Multi-Sensor Data Fusion Clustering Algorithm Based on Density Peaks.

Area of Science:

• Data Science
• Signal Processing
• Machine Learning

Background:

• Multi-sensor data fusion (MSDF) is crucial for multi-sensor target detection (MSTD).
• Clustering observations from multiple sensors presents challenges, especially without prior knowledge of clutter.
• Existing methods may struggle with constraints like cannot link (CL) and cluster size variations.

Purpose of the Study:

• To propose a novel multi-sensor clustering algorithm for MSDF in MSTD.
• To address the challenge of clustering sensor observations without prior clutter information.
• To incorporate specific constraints into the clustering process.

Main Methods:

• A novel algorithm based on the density peaks clustering (DPC) algorithm.
• Integration of the cannot link (CL) constraint, preventing data points from the same sensor in one cluster.
• Enforcement of cluster size constraints and division of overlapping clusters.

Main Results:

• The proposed algorithm demonstrates effective multi-sensor data fusion.
• Simulation results validate the algorithm's reliability in target detection scenarios.
• The algorithm successfully handles the CL constraint and cluster size limitations.

Conclusions:

• The developed multi-sensor clustering algorithm provides a robust solution for MSDF in MSTD.
• The algorithm's ability to handle complex constraints ensures accurate and reliable clustering.
• This approach enhances target detection capabilities by improving data fusion.