Maximal Privacy without Coherence

Privacy is a fundamental feature of quantum mechanics. A coherently transmitted quantum state is inherently private. Remarkably, coherent quantum communication is not a prerequisite for privacy: there are quantum channels that are too noisy to transmit any quantum information reliably that can never...

Full description

Bibliographic Details
Main Authors: Leung, Debbie W. (Author), Li, Ke (Contributor), Smith, Graeme (Author), Smolin, John A. (Author)
Other Authors: Massachusetts Institute of Technology. Center for Theoretical Physics (Contributor), Massachusetts Institute of Technology. Laboratory for Nuclear Science (Contributor)
Format: Article
Language:English
Published: American Physical Society, 2014-08-08T14:35:39Z.
Subjects:
Online Access:Get fulltext
LEADER 01746 am a22002413u 4500
001 88610
042 |a dc 
100 1 0 |a Leung, Debbie W.  |e author 
100 1 0 |a Massachusetts Institute of Technology. Center for Theoretical Physics  |e contributor 
100 1 0 |a Massachusetts Institute of Technology. Laboratory for Nuclear Science  |e contributor 
100 1 0 |a Li, Ke  |e contributor 
700 1 0 |a Li, Ke  |e author 
700 1 0 |a Smith, Graeme  |e author 
700 1 0 |a Smolin, John A.  |e author 
245 0 0 |a Maximal Privacy without Coherence 
260 |b American Physical Society,   |c 2014-08-08T14:35:39Z. 
856 |z Get fulltext  |u http://hdl.handle.net/1721.1/88610 
520 |a Privacy is a fundamental feature of quantum mechanics. A coherently transmitted quantum state is inherently private. Remarkably, coherent quantum communication is not a prerequisite for privacy: there are quantum channels that are too noisy to transmit any quantum information reliably that can nevertheless send private classical information. Here, we ask how much private classical information a channel can transmit if it has little quantum capacity. We present a class of channels N[subscript d] with input dimension d[superscript 2], quantum capacity Q(N[subscript d]) ≤ 1, and private capacity P(N[subscript d])= log d. These channels asymptotically saturate an interesting inequality P(N) ≤ (1/2)[log d[subscript A] + Q(N)] for any channel N with input dimension d[subscript A] and capture the essence of privacy stripped of the confounding influence of coherence. 
520 |a National Science Foundation (U.S.) (Grant CCF-1110961) 
520 |a National Science Foundation (U.S.) (Grant CCF-1111382) 
546 |a en 
655 7 |a Article 
773 |t Physical Review Letters