A proof-of-concept demonstration beats the best known classical protocol

Friday, October 30, 2015

Computer scientists, including Institute for Quantum Computing (IQC) members John Watrous and Richard Cleve have long been looking at protocols where quantum communication offers an advantage compared to the classical case. However technology hasn’t progressed as quickly, so researchers had previously been unable to implement the protocols.

IQC researchers Norbert Lütkenhaus, Juan Miguel Arrazola and Shihan Sajeed, in collaboration with researchers from the group of Professor Hoi-Kwong Lo in the University of Toronto, experimentally demonstrated a quantum fingerprinting system that can transmit less information than the best known classical protocol.

Quantum fingerprinting is one of the most appealing protocols in quantum communication complexity because quantum mechanics allows an exponential reduction in the transmitted information. In this problem, Alice and Bob receive inputs and based on these inputs, they send a message to a third party – the referee. The referee uses this received information to determine whether the inputs to Alice and Bob are equal, and the goal is to do this by transmitting as little information as possible.

We see a similar transmission of partial information when using software to book several people in a meeting. Everyone has their own information in their calendar, and you want an efficient protocol to find the next available time slot when everyone is free without spreading out everyone’s calendars.

Quantum fingerprinting diagram

Last year, Arrazola and Lütkenhaus proposed a practical version of a quantum fingerprinting protocol by Cleve, Watrous and collaborators. In this practical protocol, Alice and Bob’s inputs are fed to an error-correcting code that is used to modulate a sequence of very weak laser pulses. These quantum signals transmit very little information, but still allow the referee to distinguish whether the inputs are equal or not.

The research team then focused on how to implement the protocol. First, the practical laser pulse protocol was improved with a better decision rule by the referee. Additionally, a careful simulation was performed that considered experimental imperfections, and was used to design the experiment. They then developed a new error-correcting code so that they could implement the system using commercial devices. By modifying a version of a commercial Quantum Key Distribution (QKD) system using optical components, the researchers were able to transmit messages up to 100 Mbits with 66% less information than the best known classical protocol over a five-kilometre standard fibre operating at telecom wavelengths.

Quantum fingerprinting practical setup in the lab

Photo credit: Research Assistant Professor Vadim Makarov

“This doesn’t mean that the telecom industry is going to switch over to this protocol and offer that service,” said Lütkenhaus, a faculty member with the Department of Physics and Astronomy at the University of Waterloo. “It’s a first exciting step that opens the door for experimental quantum communication complexity.”

This paper, Experimental quantum fingerprinting with weak coherent pulses, published in Nature Communications on October 30, 2015, has groups approaching the team of researchers to collaborate to do more work in the field.

“For a long time, the protocols in quantum communication complexity were believed to not be implementable and so no one was working on them. This proof-of-concept experiment is a first step, but there is a lot of room for improvement,” said Arrazola, a recent PhD in the physics and astronomy department. “We are excited about the prospect of developing these results further as there’s a lot of interest in moving to new protocols.”

  1. 2019 (53)
    1. December (3)
    2. November (1)
    3. October (5)
    4. September (4)
    5. August (2)
    6. July (6)
    7. June (6)
    8. May (4)
    9. April (4)
    10. March (10)
    11. February (6)
    12. January (2)
  2. 2018 (28)
    1. October (6)
    2. September (1)
    3. August (2)
    4. July (1)
    5. June (1)
    6. May (3)
    7. April (4)
    8. March (5)
    9. February (3)
    10. January (2)
  3. 2017 (34)
  4. 2016 (44)
  5. 2015 (38)
  6. 2014 (31)
  7. 2013 (27)
  8. 2012 (33)
  9. 2011 (31)
  10. 2010 (52)
  11. 2009 (29)
  12. 2008 (31)
  13. 2007 (27)
  14. 2006 (15)
  15. 2005 (4)