Optimal Geometries for AOA Localization in the Bayesian Sense.

Area of Science:

• Signal Processing
• Estimation Theory
• Robotics

Background:

• Accurate target localization and tracking are crucial in various applications.
• Angle-of-arrival (AOA) sensing provides bearing information for localization.
• Optimal sensor placement is key to maximizing estimation accuracy.

Purpose of the Study:

• Determine optimal sensor placement for AOA target localization.
• Extend localization findings to maneuvering target tracking.
• Develop and evaluate sensor trajectory optimization algorithms.

Main Methods:

• Analytical determination of sensor locations using D- and A-optimality criteria.
• Approximation of the Bayesian Fisher Information Matrix (BFIM).
• Development of numerical search and closed-form projection algorithms for trajectory optimization.

Main Results:

• Optimal sensor placement aligns with the minor axis of the prior covariance error ellipse.
• The approximate BFIM is valid for small prior covariances.
• A novel closed-form projection algorithm outperforms numerical methods for tracking.

Conclusions:

• Sensor placement strategies for stationary targets can inform maneuvering target tracking.
• The proposed closed-form algorithm offers superior tracking performance.
• This work provides a framework for optimizing sensor configurations in dynamic environments.