Bounding quantum capacities via partial orders and complementarity
arXiv:2202.11688 · doi:10.1109/TIT.2022.3199578
Abstract
Quantum capacities are fundamental quantities that are notoriously hard to compute and can exhibit surprising properties such as superadditivity. Thus, a vast amount of literature is devoted to finding tight and computable bounds on these capacities. We add a new viewpoint by giving operationally motivated bounds on several capacities, including the quantum capacity and private capacity of a quantum channel and the one-way distillable entanglement and private key of a quantum state. These bounds are generally phrased in terms of capacity quantities involving the complementary channel or state. As a tool to obtain these bounds, we discuss partial orders on quantum channels and states, such as the less noisy and the more capable order. Our bounds help to further understand the interplay between different capacities, as they give operational limitations on superadditivity and the difference between capacities in terms of the information-theoretic properties of the complementary channel or state. They can also be used as a new approach towards numerically bounding capacities, as discussed with some examples.
25 pages, comments welcome
References in corpus (12)
- Distillation of secret key and entanglement from quantum states
- Quantum Communication With Zero-Capacity Channels
- A Sharp Fannes-type Inequality for the von Neumann Entropy
- Degenerate Quantum Codes for Pauli Channels
- The private classical capacity with a symmetric side channel and its application to quantum cryptography
- The quantum capacity with symmetric side channels
- Continuity of quantum channel capacities
- New lower bounds on the non-zero capacity of Pauli Channels
- Extensive nonadditivity of privacy
- Energy-constrained diamond norm with applications to the uniform continuity of continuous variable channel capacities
- Generic nonadditivity of quantum capacity in simple channels
- Bi-PPT channels are entanglement breaking