Signal Source Localization with Long-Term Observations in Distributed Angle-Only Sensors.

Area of Science:

• Sensor Fusion
• Target Tracking
• Signal Processing

Background:

• Angle-only sensors lack range information, necessitating sensor networks for accurate target localization.
• Tracking moving targets with sensors on moving platforms typically relies on filtering methods.

Purpose of the Study:

• To develop and evaluate novel fusion algorithms for estimating the position and velocity of moving targets using distributed, mobile angle-only sensors.
• To address the limitations of existing methods by proposing algorithms with reduced communication and computation overhead.

Main Methods:

• Gross Least Squares (LS) Algorithm: Integrates all sensor observations for joint position and velocity estimation.
• Linear LS Algorithm: Employs linear approximations for sensor/target locations and measurements to reduce data transmission.
• Truncated LS Algorithm: An extension of the linear LS algorithm that estimates target velocity via averaging.

Main Results:

• The gross LS algorithm, leveraging more observations without linear approximation, often yields better accuracy.
• The linear LS and truncated LS algorithms significantly reduce communication and computation costs.
• Performance degradation in linear LS/truncated LS algorithms can occur with prolonged data collection due to linear model mismatch.

Conclusions:

• The linear LS and truncated LS algorithms provide efficient solutions for tracking moving targets with distributed angle-only sensors, albeit with potential accuracy trade-offs.
• Algorithm selection depends on the balance between desired accuracy, communication bandwidth, and computational resources.