Research and Development

Theme

It is well known that the security of QKD is guaranteed under the assumptions of the mathematical models of users’ devices. However, it is also well recognized that there exists a gap between the mathematical assumptions and the actual operations of the devices in the real world. Another issue of the current theory is the security evaluation with finite data size. In most of the security proofs, the security is guaranteed only assuming infinitely large data length whereas in the real QKD system the data size is obviously finite. The mission of our team is to fill this gap basically from a theoretical side. Our project consists of the following subthemes:

Subtheme 1: Security analysis for finite size data
In this subtheme, we will develop the post-processing technique to guarantee the security of key distribution schemes with finite data size.
Subtheme 2: Development on efficient post-processing, privacy amplification, and error correction
In this subtheme, we will develop more efficient and calculation cost effective algorithms for privacy amplification (PA), error correction (EC) and other post-processing operations.
Subtheme 3: Device characterization and the security patch against side-channel attacks
The gap between the mathematical model and real devices could be a room for eavesdroppers to efficiently attack the system. Such kind of attack is called “side-channel attacks” and should be seriously considered in practical QKD systems. In this subtheme, first we carefully analyze the given devices both theoretically and experimentally aiming to confirm the correctness of the device model and to strengthen privacy amplification to make the scheme secure. Also we develop new device configurations or even new QKD protocols that are free from or robust against side-channels.
Subtheme 4: Security proofs for various protocols
One reason that the main protocol pursued in the Secure Photonic Network Project is based on BB84 is because it is the most advanced protocol in the sense that its security proof including realistic conditions is relatively well developed. However, it does not necessarily mean that the other protocols are not useful in practice. In this subtheme, we will develop the theory of security analyses on various proposed protocols such as B92, DPS QKD and CV QKD.

Interim results

Interim results for Team 157B-T01

Task title Outcome Date Note
1: Security analysis for finite-size data a) New theory for the finite-length security analysis of BB84 with a single photon source allowing the improvement of the key generation rate.Fig. 1a Sep. 2012 New J. Phys. 14, 093014 (2012)
b) New theory for the finite-length security analysis of the decoy methodFig. 1b Jun. 2013 QIP2013, TQC2012.
2: Development on efficient post-processing, PA, and ECFig. 2a a) New class of hash functions formulated Sep. 2012 IEEE Trans. Info. Theory, 59, 4700 (2013)
b) Fast encoding algorithm for spatially-coupled LDPC codes developed Jul. 2013 IEEE ITW2012 (Invited poster)
c) Implementation of the fast parallel error-correction decoding algorithm on GPGPU Jul. 2013 Installed on the CV QKD system developed in team 157C-T01
d) Fast and exact numerical computation algorithm for interval estimation of phase error rate (the single-photon BB84 key rate improved) Aug. 2013 QCrypt2013
3: Device characterization and the security patch against side-channel attacks a) New detector side-channel attack free (measurement device independent: MDI) QKD protocol proposedFig. 3a Mar. 2012
  • • Phys. Rev. A, 85, 042307 (2012)
  • • QCrypt2013
b) Novel optical circuits against the side-channel attack on phase modulators proposedFig. 3b May 2012 Patent applied
c) Phase correlation measurement for gain-switched laser pulses up to 1GHzFig. 3c Mar. 2013 QCrypt2013
d) Novel optical circuits compensating the detection efficiency mismatch proposed Sep. 2013 Patent applied
4: Security proofs for various protocols a) New security analysis of B92 resulting improved asymptotic key rates Jul. 2013 IEEE ISIT2013
b) Security proof of CV QKD against calibration attacks on local oscillators Aug. 2013 QCrypt2013
c) Proof of unconditional security of the DPS QKD with block-wise phase randomization Aug. 2013 QCrypt2013

Targets

Targets for Team 157B-T01

Task title Work/Milestone Due Date Note
5: System monitoring      
6: Random number generator Proof-of-principle experiment Mar. 2014  
Proto-type development Mar. 2016  
7: Efficient key distillation algorithm for CVQKD Design and implementation an efficient key distillation software for a CVQKD system. (with Gakushuin Univ. and TITECH)    
Offline demonstration Oct. 2013  
8: Document for security certificate To write a document for security certificate of QKD Mar. 2016  

Schedule

schedule

Chart 1Schedule for Team 157B-T01