2-D unitary ESPRIT-like direction-of-arrival (DOA) estimation for coherent signals with a uniform rectangular array

Area of Science:

• Signal Processing
• Array Signal Processing
• Electromagnetics

Background:

• Direction-of-arrival (DOA) estimation is crucial for applications like radar and sonar.
• Estimating coherent signals in 2-D presents computational challenges.
• Existing methods often have limitations in computational efficiency and array geometry flexibility.

Purpose of the Study:

• To propose a novel unitary transformation-based algorithm for 2-D DOA estimation of coherent signals.
• To reduce computational complexity and relax constraints on array geometry.
• To improve the decorrelation of impinging signals.

Main Methods:

• Reorganizing the covariance matrix into a block Hankel matrix for decorrelation.
• Reconstructing a new matrix to facilitate unitary transformation.
• Employing a Kronecker product of 1-D matrices for efficient 2-D unitary transformation.
• Combining ESPRIT-Like method with forward-backward averaging for enhanced decorrelation.

Main Results:

• Eigenvalue decomposition and singular value decomposition are transformed into real-valued computations, significantly reducing complexity.
• The proposed algorithm demonstrates higher computational efficiency compared to existing 2-D methods.
• The algorithm is less restrictive on array geometry.
• Simulation results validate the effectiveness of the proposed method.

Conclusions:

• The developed unitary transformation-based algorithm offers an efficient and flexible solution for 2-D DOA estimation of coherent signals.
• The method effectively reduces computational complexity through real-valued transformations.
• The approach provides a significant improvement over existing 2-D DOA estimation techniques.