Two Compensation Strategies for Optimal Estimation in Sensor Networks with Random Matrices, Time-Correlated Noises,...

Area of Science:

• Control Systems Engineering
• Signal Processing
• Networked Systems

Background:

• Signal estimation in multisensor systems is crucial but challenged by random flaws.
• Networked systems are prone to errors that degrade estimator performance.
• Existing methods struggle with time-correlated noise and packet loss.

Purpose of the Study:

• To develop robust recursive filtering and fixed-point smoothing algorithms for multisensor systems.
• To address challenges posed by random parameter matrices, time-correlated noise, deception attacks, and packet dropouts.
• To propose novel compensation strategies for enhanced estimation accuracy.

Main Methods:

• Utilized a covariance-based methodology.
• Developed two compensation strategies based on measurement prediction.
• Employed direct estimation of measurement noises and the innovation technique, overcoming limitations of measurement differencing.

Main Results:

• Designed recursive filtering and fixed-point smoothing algorithms robust to system uncertainties.
• Demonstrated the ineffectiveness of traditional measurement differencing in the presence of packet loss.
• Evaluated the performance of two proposed compensation scenarios via simulation.

Conclusions:

• The proposed methods effectively handle complex uncertainties in multisensor signal estimation.
• Direct noise estimation and innovation techniques offer a viable alternative to traditional methods.
• The study provides insights into the impact of various uncertainties on estimation accuracy.