Benchmarking and Procrustean noise reduction of entangled mixed states

Nicholas A. Peters, Tzu Chieh Wei, Paul G. Kwiat

Research output: Contribution to journalConference articlepeer-review

Abstract

Understanding quantum noise is essential for accurately creating desired quantum states and for examining a given state's evolution in any protocol. Using spontaneous parametric downconversion, we can create a wide variety of single- and two-qubit polarization states, including nearly perfect Bell states, mixed states (i.e., "noisy" states) and maximally entangled mixed states (MEMS). To characterize these states we use several different measures, including fidelity, "tangle" and linear entropy. In the course of our experiments, we have discovered and numerically investigated an extreme imbalance in the sensitivity of these different two-qubit state measures. We have also experimentally realized a "Procrustean" filtering technique to remove noise from MEMS. For moderate amounts of filtering, the experimental procedure works as desired to increase the tangle and decrease the linear entropy. However, for large amounts of filtering, the process becomes dominated by perturbations in the starting density matrix. The final outcome is a pure (i.e., zero entropy) product state (i.e., zero entanglement).

Original languageEnglish (US)
Pages (from-to)269-281
Number of pages13
JournalProceedings of SPIE - The International Society for Optical Engineering
Volume5468
DOIs
StatePublished - 2004
EventFluctuations and Noise in Photonics and Quantum Optics II - Maspalomas
Duration: May 26 2004May 28 2004

Keywords

  • Entanglement
  • MEMS
  • Procrustean filtering
  • SPDC

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Applied Mathematics
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Benchmarking and Procrustean noise reduction of entangled mixed states'. Together they form a unique fingerprint.

Cite this