Algebraic Cryptanalysis of Deterministic Symmetric Encryption

Petr Susil
EPFL, 2015
Thèse EPFL

Résumé

Deterministic symmetric encryption is widely used in many cryptographic applications. The security of deterministic block and stream ciphers is evaluated using cryptanalysis. Cryptanalysis is divided into two main categories: statistical cryptanalysis and algebraic cryptanalysis. Statistical cryptanalysis is a powerful tool for evaluating the security but it often requires a large number of plaintext/ciphertext pairs which is not always available in real life scenario. Algebraic cryptanalysis requires a smaller number of plaintext/ciphertext pairs but the attacks are often underestimated compared to statistical methods. In algebraic cryptanalysis, we consider a polynomial system representing the cipher and a solution of this system reveals the secret key used in the encryption. The contribution of this thesis is twofold. Firstly, we evaluate the performance of existing algebraic techniques with respect to number of plaintext/ciphertext pairs and their selection. We introduce a new strategy for selection of samples. We build this strategy based on cube attacks, which is a well-known technique in algebraic cryptanalysis. We use cube attacks as a fast heuristic to determine sets of plaintexts for which standard algebraic methods, such as Groebner basis techniques or SAT solvers, are more efficient. Secondly, we develop a~new technique for algebraic cryptanalysis which allows us to speed-up existing Groebner basis techniques. This is achieved by efficient finding special polynomials called mutants. Using these mutants in Groebner basis computations and SAT solvers reduces the computational cost to solve the system. Hence, both our methods are designed as tools for building polynomial system representing a cipher. Both tools can be combined and they lead to a significant speedup, even for very simple algebraic solvers.

Official source

https://infoscience.epfl.ch/record/210605?ln=fr

À propos de ce résultat

Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.

Concepts associés

Chargement

Publications associées

Chargement

Petr Susil
EPFL, 2015
Thèse EPFL

Résumé

Official source

https://infoscience.epfl.ch/record/210605?ln=fr

À propos de ce résultat

Concepts associés (15)

Concepts associés

Chargement

Publications associées (10)

Chargement

Afficher plus

Publications associées

Chargement

Publications associées (10)

Analysis of lightweight stream ciphers

Simon Fischer

Stream ciphers are fast cryptographic primitives to provide confidentiality of electronically transmitted data. They can be very suitable in environments with restricted resources, such as mobile devices or embedded systems. Practical examples are cell phones, RFID transponders, smart cards or devices in sensor networks. Besides efficiency, security is the most important property of a stream cipher. In this thesis, we address cryptanalysis of modern lightweight stream ciphers. We derive and improve cryptanalytic methods for different building blocks and present dedicated attacks on specific proposals, including some eSTREAM candidates. As a result, we elaborate on the design criteria for the development of secure and efficient stream ciphers. The best-known building block is the linear feedback shift register (LFSR), which can be combined with a nonlinear Boolean output function. A powerful type of attacks against LFSR-based stream ciphers are the recent algebraic attacks, these exploit the specific structure by deriving low degree equations for recovering the secret key. We efficiently determine the immunity of existing and newly constructed Boolean functions against fast algebraic attacks. The concept of algebraic immunity is then generalized by investigating the augmented function of the stream cipher. As an application of this framework, we improve the cryptanalysis of a well-known stream cipher with irregularly clocked LFSR's. Algebraic attacks can be avoided by substituting the LFSR with a suitable nonlinear driving device, such as a feedback shift register with carry (FCSR) or the recently proposed class of T-functions. We investigate both replacement schemes in view of their security, and devise different practical attacks (including linear attacks) on a number of specific proposals based on T-functions. Another efficient method to amplify the nonlinear behavior is to use a round-based filter function, where each round consists of simple nonlinear operations. We use differential methods to break a reduced-round version of eSTREAM candidate Salsa20. Similar methods can be used to break a related compression function with a reduced number of rounds. Finally, we investigate the algebraic structure of the initialization function of stream ciphers and provide a framework for key recovery attacks. As an application, a key recovery attack on simplified versions of eSTREAM candidates Trivium and Grain-128 is given.

EPFL2008

Chargement

Symmetric cryptographic primitives such as block and stream ciphers are the building blocks in many cryptographic protocols. Having such blocks which provide provable security against various types of attacks is often hard. On the other hand, if possible, such designs are often too costly to be implemented and are usually ignored by practitioners. Moreover, in RFID protocols or sensor networks, we need lightweight and ultra-lightweight algorithms. Hence, cryptographers often search for a fair trade-off between security and usability depending on the application. Contrary to public key primitives, which are often based on some hard problems, security in symmetric key is often based on some heuristic assumptions. Often, the researchers in this area argue that the security is based on the confidence level the community has in their design. Consequently, everyday symmetric protocols appear in the literature and stay secure until someone breaks them. In this thesis, we evaluate the security of multiple symmetric primitives against statistical and algebraic attacks. This thesis is composed of two distinct parts: In the first part, we investigate the security of RC4 stream cipher against statistical attacks. We focus on its applications in WEP and WPA protocols. We revisit the previous attacks on RC4 and optimize them. In fact, we propose a framework on how to deal with a pool of biases for RC4 in an optimized manner. During this work, we found multiple new weaknesses in the corresponding applications. We show that the current best attack on WEP can still be improved. We compare our results with the state of the art implementation of the WEP attack on Aircrack-ng program and improve its success rate. Next, we propose a theoretical key recovery and distinguishing attacks on WPA, which cryptographically break the protocol. We perform an extreme amount of experiments to make sure that the proposed theory matches the experiments. Finally, we propose a concrete theoretical and empirical proof of all our claims. These are currently the best known attacks on WEP and WPA. In the second part, we shed some lights on the theory behind ElimLin, which is an algorithm for solving multivariate polynomial systems of equations. We attack PRESENT and LBlock block ciphers with ElimLin algorithm and compare their security using this algebraic technique. Then, we investigate the security of KATAN family of block ciphers and multi-purpose cryptographic primitive ARMADILLO against algebraic attacks. We break reduced-round versions of several members of KATAN family by proposing a novel pre-processing technique on the original algebraic representation of the cipher before feeding it to a SAT solver. Finally, we propose a devastating practical key recovery attack against the ARMADILLO1 protocol, which breaks it in polynomial time using a few challenge-response pairs.

EPFL2012

Chargement

Neutrality-Based Symmetric Cryptanalysis

Shahram Khazaei

Cryptographic primitives are the basic components of any cryptographic tool. Block ciphers, stream ciphers and hash functions are the fundamental primitives of symmetric cryptography. In symmetric cryptography, the communicating parties perform essentially the same operation and use the same key, if any. This thesis concerns cryptanalysis of stream ciphers and hash functions. The main contribution of this work is introducing the concept of probabilistic neutrality for the arguments of a function, a generalization of the definition of neutrality. An input argument of a given function is called neutral if it does not affect the output of the function. This simple idea has already been implicitly used in key recovery cryptanalysis of block ciphers and stream ciphers. However, in 2004, Biham and Chen explicitly used the idea of neutrality to speed up collision finding algorithms for hash functions. We call an input argument of a function probabilistic neutral if it does not have a "significant" influence on the output of the function. Simply stated, it means that if the input argument is changed, the output of the function stays the same with a probability "close" to one. We will exploit the idea of probabilistic neutrality to assess the security of several stream ciphers and hash functions. Interestingly, all our cryptanalyses rely on neutrality and/or probabilistic neutrality. In other words, these concepts will appear as a common ingredient in all of our cryptanalytic algorithms. To the best of our knowledge, this is the first time that the probabilistic neutrality has found diverse applications in cryptanalysis.

EPFL2010

Chargement

Afficher plus

Analysis of lightweight stream ciphers

Simon Fischer

EPFL2008

EPFL2012