A New Statistical Method for Determining the Clutter Covariance Matrix in Spatial-Temporal Adaptive Processing of a...

Area of Science:

• Radar Signal Processing
• Statistical Signal Processing
• Array Signal Processing

Background:

• Accurate clutter covariance matrix estimation is crucial for Space-Time Adaptive Processing (STAP) in radar systems.
• Existing statistical methods, like SMI based on QR distribution, have limitations, especially in heterogeneous environments.
• Multiple-Input-Multiple-Output (MIMO) radar with Time Division Multiplexing (TDM) presents unique challenges for clutter estimation.

Purpose of the Study:

• To introduce and analyze a novel statistical method for clutter covariance matrix estimation in STAP.
• To compare the proposed method with existing statistical and non-statistical approaches.
• To validate the method's performance for MIMO radar with TDM.

Main Methods:

• Development of a new statistical estimation method based on LU distribution with partial pivoting.
• Presentation of the STAP algorithm for the standard statistical Sample Matrix Inversion (SMI) method using QR distribution.
• Extensive analysis comparing statistical and non-statistical clutter covariance matrix estimation techniques.

Main Results:

• Simulation results confirm the validity of the proposed LU distribution-based model and theoretical assumptions.
• The new method demonstrated more accurate object detection compared to other statistical methods in specific computational examples.
• The study highlights a gap in current research concerning non-statistical methods for heterogeneous environments.

Conclusions:

• The proposed LU distribution-based statistical method provides a more accurate approach to clutter covariance matrix estimation in STAP for MIMO radar.
• This research addresses the growing global focus on non-statistical methods by offering a novel statistical alternative.
• The findings contribute to advancing STAP performance, particularly in challenging heterogeneous environments.