An Optimal Linear Fusion Estimation Algorithm of Reduced Dimension for T-Proper Systems with Multiple Packet Dropouts

Area of Science:

• Signal Processing
• Control Systems Engineering
• Information Theory

Background:

• Multi-sensor systems are crucial for enhanced data acquisition and decision-making.
• Packet dropouts and correlated noises in data transmission degrade system performance.
• Centralized fusion linear estimation is vital for integrating information from multiple sensors.

Purpose of the Study:

• To develop an optimal linear fusion filtering algorithm for multi-sensor systems with packet dropouts and correlated noises.
• To reduce the computational complexity of state estimation in such systems.
• To leverage the tessarine domain for efficient data processing.

Main Methods:

• Modeling packet dropouts using independent Bernoulli distributed random variables.
• Applying T1 and T2-properness conditions in the tessarine domain.
• Developing a least-mean-squares optimal linear fusion filtering algorithm.

Main Results:

• Achieved a reduction in problem dimension and computational cost.
• Proposed an optimal linear fusion filtering algorithm in the tessarine domain.
• Demonstrated superior performance and computational advantages over conventional real-field methods through simulations.

Conclusions:

• The proposed tessarine-domain methodology offers significant computational savings for centralized fusion linear estimation.
• The developed algorithm provides an optimal (in the least-mean-squares sense) solution for state estimation in multi-sensor systems with packet dropouts.
• Simulation results validate the effectiveness and practical applicability of the proposed approach.