Locally Minimum-Variance Filtering of 2-D Systems Over Sensor Networks With Measurement Degradations: A Distributed...

Area of Science:

• Control Systems Engineering
• Networked Systems
• Signal Processing

Background:

• Recursive filtering is crucial for state estimation in dynamic systems.
• Sensor networks introduce complexities like measurement degradation and coupling perturbations.
• Distributed filtering is essential for cooperative estimation in large-scale networks.

Purpose of the Study:

• To design distributed recursive filters for 2-D systems in sensor networks.
• To minimize the upper bound on the second-order moment of the filtering error.
• To develop a scalable filtering algorithm for networked systems.

Main Methods:

• Modeling measurement degradations and stochastic coupling perturbations.
• Utilizing stochastic analysis and induction for error variance upper bound construction.
• Determining filter gain parameters based on network sparsity.

Main Results:

• A distributed recursive filtering strategy is proposed for 2-D systems.
• The method ensures a locally minimal upper bound on the filtering error variance.
• The developed algorithm is proven to be scalable for sensor networks.

Conclusions:

• The proposed filtering strategy effectively addresses recursive filtering problems in sensor networks.
• The approach provides a robust method for state estimation under realistic network conditions.
• The scalability of the algorithm is validated, making it suitable for practical applications.