Abstract:Quantum relative entropy programs are convex optimization problems which minimize a linear functional over an affine section of the epigraph of the quantum relative entropy function. Recently, the self-concordance of a natural barrier function was proved for this set. This has opened up the opportunity to use interior-point methods for nonsymmetric cone programs to solve these optimization problems. In this paper, we show how common structures arising from applications in quantum information theory can be exploited to improve the efficiency of solving quantum relative entropy programs using interior-point methods. First, we show that the natural barrier function for the epigraph of the quantum relative entropy composed with positive linear operators is optimally self-concordant, even when these linear operators map to singular matrices. Compared to modelling problems using the full quantum relative entropy cone, this allows us to remove redundant log determinant expressions from the barrier function and reduce the overall barrier parameter. Second, we show how certain slices of the quantum relative entropy cone exhibit useful properties which should be exploited whenever possible to perform certain key steps of interior-point methods more efficiently. We demonstrate how these methods can be applied to applications in quantum information theory, including quantifying quantum key rates, quantum rate-distortion functions, quantum channel capacities, and the ground state energy of Hamiltonians. Our numerical results show that these techniques improve computation times by up to several orders of magnitude, and allow previously intractable problems to be solved.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the structure utilization problem in the quantum relative entropy program in order to improve the efficiency of solving these optimization problems using the interior - point method. Specifically, the paper focuses on how to improve the key steps of the interior - point method for solving the quantum relative entropy program by utilizing some common structures in quantum information theory, thereby reducing the computation time and memory requirements.
### Background and Problem Description of the Paper
Quantum Relative Entropy (QRE) is an important function in quantum information theory, which is used to measure the difference between two quantum states. It is defined as:
\[ S(X \parallel Y) := \text{tr}[X(\log(X) - \log(Y))] \]
On the domain \(\{(X, Y) \in \mathcal{H}_+^n \times \mathcal{H}_+^n : X \ll Y\}\), where \(X \ll Y\) means \(\text{ker}(Y) \subseteq \text{ker}(X)\), \(\log\) represents the matrix logarithm, and \(\mathcal{H}_+^n\) represents the set of positive semi - definite Hermitian matrices.
The quantum relative entropy program is a class of convex optimization problems, with the objective of minimizing the quantum relative entropy under linear constraints. This type of problem can be expressed as:
\[ \min_{X \in \mathcal{H}_+^n} S(G(X) \parallel H(X)), \quad \text{subj. to } A(X) = b, \quad X \succeq 0 \]
where \(G: \mathcal{H}_+^n \to \mathcal{H}_+^m\) and \(H: \mathcal{H}_+^n \to \mathcal{H}_+^m\) are positive linear maps, and \(A: \mathcal{H}_+^n \to \mathbb{R}^p\) and \(b \in \mathbb{R}^p\) encode the linear constraints.
### Main Contributions of the Paper
1. **Optimized Barrier Function**: The paper proposes a new barrier function that is more efficient in dealing with the quantum relative entropy program. Specifically, for slices of the form \((t, X) \in \mathbb{R} \times \mathcal{H}_+^n\), the barrier function is defined as:
\[ (t, X) \in \mathbb{R} \times \mathcal{H}_+^n \mapsto -\log(t - S(G(X) \parallel H(X))) - \log \det(X) \]
This barrier function is an \((n + 1)\)-logarithmically homogeneous self - concordant barrier and is optimal.
2. **Structure Utilization**: The paper shows how to utilize the common structures in \(G\) and \(H\) (such as block - diagonal, low - rank, and quantum conditional entropy structures) to simplify the process of solving the Newton equation. These techniques can significantly reduce the computation time and memory requirements, making previously intractable problems feasible.
### Application Examples
The paper shows several applications of these techniques in quantum information theory, including:
- **Quantum Key Rate**: Calculate the key rate in the quantum key distribution protocol.
- **Quantum Rate - Distortion Function**: Quantify the degree to which a quantum source can be compressed without exceeding the maximum allowed distortion.
- **Quantum Channel Capacity**: Calculate the maximum transmission rate of a quantum channel.
- **Ground - State Energy of Hamiltonian**: Calculate the ground - state energy of a local Hamiltonian.
### Experimental Results
The experimental results show that the method proposed in the paper can reduce the computation time by several orders of magnitude and can solve large - scale problems that were previously unmanageable.
In conclusion, through optimizing the barrier function and utilizing structures, this paper significantly improves the efficiency of solving the quantum relative entropy program and provides effective solutions for several important problems in quantum information theory.