A New Transient Attack on the Kish Key Distribution System

Abstract

The Kish Key Distribution (KKD) system has been proposed as a classical alternative to quantum key distribution, making use of temperature-matched thermal noise. Previous analyses assume instant propagation of signals along the cable connecting the two users. We describe a new attack that takes an advantage of propagation delays. At the start of each bit period, the noise temperature will then be increased from zero to its final value. During this process, the noise temperature variation will take time to propagate along the line, resulting in a temperature mismatch. We analyze the information leak due to this effect and consider several potential mitigation schemes.

Comments on “A New Transient Attack on the Kish Key Distribution System”

“…We have confirmed GAA's conclusion [15] that their attack works with about p = 0.7-0.8, where p is Eve's probability of successfully guessing the key-bits. These values of p require four stages of the simplest XORbased privacy amplification in order to reduce p to its ideal range of 0.5 < p < 0.5006 [33], which implies a corresponding 16-fold slowdown of the key exchange.…”

“…The present paper concerns the classical statistical-physicsbased Kirchhoff-law-Johnson-noise (KLJN) key distribution system, delineated in Figure 1, which is no exception to the tradition of the research area, and the creation of the KLJN schemes [3,4] immediately triggered attacks [5][6][7]. The various attacks [5][6][7][8][9][10][11][12][13][14][15][16] have led to useful discussions [17][18][19][20][21][22][23], including corrections of flaws in the attacks [19][20][21][22][23] and developments of new defense protocols [5,10,11,13,24,25] as well as protocols that have increased immunity against attacks in general [24][25][26][27]. Furthermore, KLJN schemes that are totally immune to a certain attack have been presented [13,[28][29][30] as has a new system that is immune to all existing attacks [31].…”

“…Recently, Gunn, Allison and Abbott (GAA) published an interesting paper [15] with the first attack utilizing transients at the beginning of the bit-exchange. Their idea is impressively simple and involves monitoring the mean-square voltage before the front of the transient reaches the other end of the communication cable.…”

“…The mean-square generator voltages and the resistance values are independent random variables, and hence Eve cannot relate the measured transient mean-square voltage to the resistance value. Thus GAA's transient attack [15] yields zero information leak about the key (p = 0.5).…”

A Static-Loop-Current Attack against the KLJN Secure Key Exchange System

“…None of them could impair the unconditional security of the scheme, since either there was a proper defense scheme against or attack was based on misconceptions [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][27][28][29][30], including experimental errors. On the other hand, passive attacks against practical KLJN systems deviating from the ideal assumptions of the scheme [2] were discussed in several papers [4][5][6][7][8][9][10][11][12][13]. For example, parasitic or non-ideal features such as transients, wire's resistance, cable capacitance, temperature differences, delay effects and transients, etc.…”