Paper ID | D3-S3-T4.2 |
Paper Title |
Discrimination of quantum states under locality constraints in the many-copy setting |
Authors |
Hao-Chung Cheng, National Taiwan University, Taiwan; Andreas Winter, Universitat Autònoma de Barcelona, Spain; Nengkun Yu, University of Technology Sydney, Australia |
Session |
D3-S3-T4: Quantum Detection & Estimation |
Chaired Session: |
Wednesday, 14 July, 22:40 - 23:00 |
Engagement Session: |
Wednesday, 14 July, 23:00 - 23:20 |
Abstract |
We study the discrimination of a pair of orthogonal quantum states in the many-copy setting. This is not a problem when arbitrary quantum measurements are allowed, as then the states can be distinguished perfectly even with one copy. However, it becomes highly nontrivial when we consider states of a multipartite system and locality constraints are imposed. We hence focus on the restricted families of measurements such as local operation and classical communication (LOCC), separable operations (SEP), and the positive-partial-transpose operations (PPT) in this paper. We first study asymptotic discrimination of an arbitrary multipartite entangled pure state against its orthogonal complement using LOCC/SEP/PPT measurements. We prove that the incurred optimal average error probability always decays exponentially in the number of copies, by proving upper and lower bounds on the exponent. In the special case of discriminating a maximally entangled state against its orthogonal complement, we determine the explicit expression for the optimal average error probability, thus establishing the associated Chernoff exponent. Our technique is based on the idea of using PPT operations to approximate LOCC. Then, we show an infinite separation between SEP and PPT operations by providing a pair of states constructed from an unextendible product basis (UPB): they can be distinguished perfectly by PPT measurements, while the optimal error probability using SEP measurements admits an exponential lower bound. On the technical side, we prove this result by providing a quantitative version of the well-known statement that the tensor product of UPBs is UPB.
|