Abstract:Convex optimization methods have been extensively used in the fields of communications and signal processing. However, the theory of quaternion optimization is currently not as fully developed and systematic as that of complex and real optimization. To this end, we establish an essential theory of convex quaternion optimization for signal processing based on the generalized Hamilton-real (GHR) calculus. This is achieved in a way which conforms with traditional complex and real optimization theory. For rigorous, We present five discriminant theorems for convex quaternion functions, and four discriminant criteria for strongly convex quaternion functions. Furthermore, we provide a fundamental theorem for the optimality of convex quaternion optimization problems, and demonstrate its utility through three applications in quaternion signal processing. These results provide a solid theoretical foundation for convex quaternion optimization and open avenues for further developments in signal processing applications.

What problem does this paper attempt to address?

The paper attempts to address the problem of establishing a convex quaternion optimization theory in the field of signal processing. Specifically, while the real-valued and complex-valued convex optimization theories are already very mature and widely applied in areas such as communications, machine learning, and signal processing, the quaternion optimization theory is currently not sufficiently developed and systematized. To bridge this gap, the paper establishes the fundamental theory of convex optimization in quaternion signal processing through the generalized Hamilton-real (GHR) calculus method. The main contributions of the paper include: 1. **Five Discriminant Theorems**: Through GHR calculus, the paper establishes five discriminant theorems for convex functions in the quaternion domain, covering gradient monotonicity and second-order characteristics. 2. **Definition and Criteria of Strongly Convex Functions**: The paper proposes a clear definition of strongly convex functions in the quaternion domain and provides four discriminant criteria, which are consistent with the theorems of real-valued and complex-valued strongly convex functions. 3. **Fundamental Theorem of Optimality and Applications**: The paper presents a fundamental theorem on the optimality of convex quaternion problems and provides three practical application cases in the field of signal processing, including quaternion linear mean square error filters, quaternion projection onto affine equality constraints, and quaternion minimum variance beamforming. These results provide a solid theoretical foundation for convex quaternion optimization and open up new avenues for subsequent developments in signal processing applications.

Convex Quaternion Optimization for Signal Processing: Theory and Applications

A general framework for constrained convex quaternion optimization

Quaternion Filtering Based on Quaternion Involutions and Its Application in Signal Processing

A Neurodynamic Approach to Nonsmooth Quaternion Distributed Convex Optimization With Inequality and Affine Equality Constraints

Standard Dual Quaternion Optimization and Its Applications in Hand-Eye Calibration and SLAM

Convolution theorems associated with quaternion linear canonical transform and applications

The HR-Calculus: Enabling Information Processing with Quaternion Algebra

Randomized Quaternion Matrix UTV Decomposition and Its Applications in Quaternion Matrix Optimization

Multilinear analysis of quaternion arrays: theory and computation

On generalized filtering methods for quaternion-based orientation estimation

Augmented Quaternion and Augmented Unit Quaternion Optimization

Efficient quaternion CUR method for low-rank approximation to quaternion matrix

CUR and Generalized CUR Decompositions of Quaternion Matrices and their Applications

Penalty Method for Constrained Distributed Quaternion-Variable Optimization

Properties and applications of quaternion quadratic phase Fourier transforms

A hybrid algorithm for quadratically constrained quadratic optimization problems

Flexible Quaternion Generalized Minimal Residual Method for Ill-Posed Quaternion Inverse Problems

An Optimisation Approach to Non-Separable Quaternion-Valued Wavelet Constructions

Nonconvex Robust Quaternion Matrix Completion for Imaging Processing