FLUCTUATION AND NOISE EXPLOITATION LABORATORY
Dept. of Electrical and Computer Engineering, Texas A&M University
secure key distribution
Unconditionally (information theoretically) secure key exchange scheme based on enhanced Johnson-noise, the Second Law of Thermodynamics and Kirchhoff's Law.
Latest hot news:
The last updates below are from November 14, 2014. A lot has happened since but no time for an update. It will be done during Summer/2016 when the whole structure of this page will be revised and new pages added. Until then, check the Publication List (sections C-D-E) for news: click on
Up to November 14, 2014:
- We identified the experimental error leading to the giant information leak in the Gunn-Allison-Abbott (GAA) attack, see the next paragraph below. It is a fundamental error of engineering design. For the huge information leak results, they use a commercial cable attenuator, which is a voltage divider (T-element) breaking the single Kirchhoff-loop of the KLJN system into two coupled loops. Thus, what they measured was not a KLJN system. Another flaw in a Nature-based journal! Note: a fully built KLJN would not have been able to function due its built-in current/voltage comparison units at the two ends. Here is the link to the pdf file of our critical paper with these details (accepted for publication). click here.
- The Gunn-Allison-Abbott (GAA) attack is published in (Nature) Scientific Reports. They also tested our breakthrough solution below (see the Entropy paper) and found that it protected even against their attack, see Equation (19) and the text around it. Note, they incorrectly say that alpha must be measured because all the resistors and the line parameters are public information in the KLJN system, and Alice and Bob knows them anyway. Concerning the rest of the GAA paper, all our former objections against the GAA attack holds, see those papers below. To read their Nature paper, click here.
- Breakthrough: The famous old (Bergou)-Scheuer-Yariv wire resistance attack (for its correct treatment, see paper 13 below) is totally eliminated now. The new defense method, which is to increase the noise temperature of the smaller resistances, nullifies not only the (Bergou)-Scheuer-Yariv attack, but totally eliminates also the new, more efficient Second Law attack. Accepted for publication in Entropy. Click here for the paper.
- Zoltan Gingl and Robert Mingesz gave a mathematical security proof showing that only Gaussian noises can be applied in the KLJN system, like it is in the original Johnson noise based scheme (resistor thermal noise without extra generators). PloS ONE (2014 April), Click here to read the paper
- Criticism-2 of the GAA attack paper. Due to the incorrect math/physics foundations of the Gunn-Allison-Abbott (GAA) attack system, we carried out the correct analysis proving that it does not offer anything more efficient than the old comparison between the mean-square voltages at the cable ends, see the Kish-Scheuer paper below, in this page. We identified some of the possible experimental artifacts in GAA's work that could lead to the unphysical results. Our treatment shows that parasitic deterministic currents in the loop, non-Gaussianity, aliasing effects, non-linearities, etc, can affect the security practical KLJN systems if their design is poor. This paper is published. Click here to read the paper.
- Criticism-1 of the GAA attack paper. Proof that, in accordance with several laws of physics, no waves exist in a short cable at low frequencies. Note, this discussion is mostly irrelevant for security however concludes a historical open question. The paper was motivated by the opposite claim in the GAA attack manuscript. This paper is published. Click here to read the paper.
- Our response to the Bennett-Riedel manuscript is published in PLOS ONE, click here
- The most general security proof based on the foundation of classical physics: the continuity of functions describing stable classical physical systems. With Claes Granqvist, published in Quantum Information Processing as invited paper, click here for the preprint and find the journal version by clicking on the DOI link there.
- Yessica: Efficient reduction of the bit error probability (10^-12 or less at practical conditions) with combined current and voltage monitoring, accepted for publication in Journal of Computational Electronics, click here
- Yessica: Error probability decays exponentially with increasing bandwidth or bit exchange duration, accepted for publication in PLOS ONE, click here
- Elias: Unconditionally secure Smart Grid (electrical power distribution), PLOS ONE, 2013, click here
- Physical Uncloneable Function (PUF) Non-Counterfeitable Hardware Keys, FNL, in press, click here
- April 16, 2013. After one year that we had requested it, the Kish cypher Wikipedia page has been deleted,
see our supporting comments at the Wikipedia Talk page, which finally triggered the deletion action. click here
- Seven new KLJN systems with strongly enhanced security and transient protocol for the non-ideal situations click here
Wikipedia disclaimer: Up to April 16th, 2013, there was a Wikipedia site called "Kish cypher", where even the name "Kish cypher" was incorrect because this scheme is not a cypher but a secure key exchange protocol built on a specific physical system and its laws. That wiki page got fortunately deleted but its various versions are still circulating on the web and in wiki-based books. It was apparent that many of the contributors either have not read the papers or have not been able to follow them. Inspired by these events, a book shall be completed and published by World Scientific (expected at the end of 2014):
Relevant: on August 26, 2013, Scholarpedia published a brief introduction on KLJN click here
Image of the simplified scheme in the Swedish magazine Ny Teknik (New Technology)
click to see/enlarge:
The fully protected KLJN scheme: