Lattice Stuff

We — with Jean-Charles FaugèreRobert Fitzpatrick and Ludovic Perret – managed to finish our work on the cryptanalysis of all proposed parameters of the public-key encryption scheme proposed at PKC 2012 by Huang, Liu and Yang. The key observation is that the scheme can be viewed as an easy LWE instance:

In this paper, we investigate the security of a public-key encryption scheme introduced by Huang, Liu and Yang (HLY) at PKC’12. This new scheme can be provably reduced to the hardness of solving a set of quadratic equations whose coefficients of highest degree are chosen according to a discrete Gaussian distributions. The other terms being chosen uniformly at random. Such a problem is a variant of the classical problem of solving a system of non-linear equations (PoSSo), which is known to be hard for random systems. The main hypothesis of Huang, Liu and Yang is that their variant is not easier than solving PoSSo for random instances. In this paper, we disprove this hypothesis. To this end, we exploit the fact that the new problem proposed by Huang, Liu and Yang reduces to an easy instance of the Learning With Errors (LWE) problem. The main contribution of this paper is to show that security and efficiency are essentially incompatible for the HLY proposal. That is, one cannot find parameters which yield a secure and a practical scheme. For instance, we estimate that a public-key of at least 1.03 GB is required to achieve 80-bit security against known attacks. As a proof of concept, we present practical attacks against all the parameters proposed Huang, Liu and Yang. We have been able to recover the private-key in roughly one day for the first challenge proposed by HLY and in roughly three days for the second challenge.

Furthermore, I gave a talk yesterday on solving LWE with binary secret using a variant of the BKW algorithm at SIAM AG’13.


BKW: Update

We have updated our pre-print titled “On the Complexity of the BKW Algorithm on LWE” on ePrint.

There are two main changes and the reasons why I am mentioning this update here.

  1. We included a more thorough comparison with other approaches, in particular, with lattice reduction (reducing LWE to SIS). To our surprise, BKW is quite competitive even in relatively modest dimensions. For Regev’s and Lindner-Peikert’s parameter sets (as interpreted here) we get that BKW is at least as fast as BKZ starting in dimension n \approx 250, which I find very low (see Table 4 on page 19).
  2. We also provide an alternative approximating for the running time of BKZ. The standard estimate due to Lindner-Peikert is \log_2 T_{sec} = \log_2 1.8/\delta_0 - 110 where \delta_0 is the targeted root hermit factor. Interpolating estimates from the BKZ 2.0 simulator and reflecting on the doubly exponential running time of BKZ in the blocksize \beta we found: \log_2 T_{sec} = \log_2 0.009/\delta^2_0 - 27. However, since this might be controversial, we include estimates for both models.


I am currently attending ESC 2013 in Mondorf, Luxembourg. Over dinner someone mentioned that there is no known reduction from LPN to lattice reduction, i.e., it is not known that you can solve LPN with LLL and friends.  This seems rather strange to me, because the standard lattice attack on LWE seems to be carrying over as is:

sage: n = 100 # number of variables
sage: m = 400 # number of samples
sage: A = random_matrix(GF(2), m, n)
sage: s = random_vector(GF(2), n) # our secret
sage: p = 0.25 # our error rate

sage:  v = A*s + vector(GF(2),[1 if random() < p else 0 for _ in range(m)])

# we are searching for a short vector in the dual lattice
sage: B = A.kernel().matrix()
sage: L = B.change_ring(ZZ).LLL()

# because a short vector there, means few additions which means a higher bias in the sum
sage: Av = A.augment(v)
sage: sum(map(lambda x: abs(x) % 2,L[0])), (L[0]*Av)[-1]

Of course, this means running lattice reduction many times, but still: what am I missing?

PS: Obligatory, Sage cell here.

On the Complexity of the BKW Algorithm on LWE

We (with Carlos CidJean-Charles FaugèreRobert Fitzpatrick and Ludovic Perret) have finally managed to put our work on BKW on ePrint.


In this paper we present a study of the complexity of the Blum-Kalai-Wasserman (BKW) algorithm when applied to the Learning with Errors (LWE) problem, by providing refined estimates for the data and computational effort requirements for solving concrete instances of the LWE problem. We apply this refined analysis to suggested parameters for various LWE-based cryptographic schemes from the literature and, as a result,  provide new upper bounds for the concrete hardness of these LWE-based schemes.

The source code of our (not very efficient!) implementation of BKW is available on bitbucket.

An All-In-One Approach to Differential Cryptanalysis for Small Block Ciphers

a paper that I wrote with Gregor Leander is finally done, out and accepted for presentation at SAC.

We present a framework that unifies several standard differential techniques. This unified view allows us to consider many, potentially all, output differences for a given input difference and to combine the information derived from them in an optimal way. We then propose a new attack that implicitly mounts several standard, truncated, impossible, improbable and possible future variants of differential attacks in parallel and hence allows to significantly improve upon known differential attacks using the same input difference. To demonstrate the viability of our techniques, we apply them to KATAN-32. In particular, our attack allows us to break 115 rounds of KATAN-32, which is 37 rounds more than previous work. For this, our attack exploits the non-uniformity of the difference distribution after 91 rounds which is 20 rounds more than the previously best known differential characteristic. Since our results still cover less than 1/2 of the cipher, they further strengthen our confidence in KATAN-32’s resistance against differential attacks.


SAT Solvers for Sage

One of the most efficient techniques for solving polynomial systems over \mathbb{F}_2 is to convert the problem to a satisfiability problem and to use a standard SAT solver. In the past, I have used CryptoMiniSat and either my own ANF to CNF converter scripts based on Gregory Bard’s ideas or PolyBoRi’s script.

However, this setup leaves much to be desired:

  1. It’s all based on string parsing which has some overhead.
  2. Usually the instances produced using PolyBoRi’s conversion method are faster to solve. However, as the number of variables per equation increase this method becomes essentially exponentially more expensive. Hence, a compromise between the two techniques is needed.
  3. We don’t have access to learnt clauses and conflict clauses.
  4. It all feels a bit duct taped and fragile, partly because the code is not shipped with Sage.

At #418 I just finished a much nicer interface to various SAT solvers. Here are some features. Continue reading “SAT Solvers for Sage”

Summer School on Tools :: Mykonos, Greece :: 28.5 – 1.6.

Slightly redacted announcement for the 2012 Summer School on Tools below.

Following the success of the ECRYPT Workshop on Tools for Cryptanalysis 2010,the ECRYPT II Symmetric Techniques Virtual Lab (SymLab) is pleased to announce the 2012 Summer School on Tools. Covering selected topics in both symmetric and asymmetric cryptography, this summer school will provide a thorough overview of some of the most important cryptographic tools that emerged in recent years. While the summer school is aimed primarily at postgraduate students, attendance is open to all. Continue reading “Summer School on Tools :: Mykonos, Greece :: 28.5 – 1.6.”