Misuse Attacks on Post-quantum Cryptosystems

Many post-quantum cryptosystems which have been proposed in the National Institute of Standards and Technology (NISI) standardization process follow the same meta-algorithm, but in different algebras or different encoding methods. They usually propose two constructions, one being weaker and the other requiring a random oracle. We focus on the weak version of nine submissions to NISI. Submitters claim no security when the secret key is used several times. In this paper, we analyze how easy it is to run a key recovery under multiple key reuse. We mount a classical key recovery under plaintext checking attacks (i.e., with a plaintext checking oracle saying if a given ciphertext decrypts well to a given plaintext) and a quantum key recovery under chosen ciphertext attacks. In the latter case, we assume quantum access to the decryption oracle.

Resistance against Adaptive Plaintext-Ciphertext Iterated Distinguishers

Asli Bay, Atefeh Mashatan, Serge Vaudenay

Decorrelation Theory deals with general adversaries who are mounting iterated attacks, i.e., attacks in which an adversary is allowed to make d queries in each iteration with the aim of distinguishing a random cipher C from the ideal random cipher C^*. A bound for a non-adaptive iterated distinguisher of order d, who is making plaintext (resp. ciphertext) queries, against a 2d-decorrelated cipher has already been derived by Vaudenay at EUROCRYPT '99. He showed that a 2d-decorrelated cipher resists against iterated non-adaptive distinguishers of order d when iterations have almost no common queries. More recently, Bay et al. settled two open problems arising from Vaudenay's work at CRYPTO '12, yet they only consider non-adaptive iterated attacks. Hence, a bound for an adaptive iterated adversary of order d, who can make both plaintext and ciphertext queries, against a 2d-decorrelated cipher has not been studied yet. In this work, we study the resistance against this distinguisher and we prove the bound for an adversary who is making adaptive plaintext and ciphertext queries depending on the previous queries to an oracle.

2012

Analysis of the BIKE post-quantum cryptographic protocols and the Legendre pseudorandom function

Dusan Kostic

The field of post-quantum cryptography studies cryptographic systems that are secure against an adversary in possession of a quantum computer. In 2017, the National Institute of Standards and Technology (NIST) initiated a process to standardize quantum-resistant public-key cryptographic algorithms (NIST PQC Project). In this thesis we analyze the performance and security of the Bit-Flipping Key Encapsulation Mechanism (BIKE) -- one of the candidates in the NIST PQC project which advanced to the second round of the standardization process. BIKE is a code-based cryptographic system featuring three different variants of the protocol. In the first round of the NIST PQC project BIKE offered security only against chosen-plaintext attacks (CPA). In the second round, BIKE introduced three new variants that are claimed to be secure also against chosen-ciphertext attacks (CCA). Firstly, we build a secure implementation of the CCA protocol and show that its performance characteristics are only negligibly worse than the CPA variant. In the key decapsulation phase of the protocol BIKE uses a decoding algorithm which fails with some probability, called the Decoding Failure Rate (DFR). We analyze the DFR of two decoders used in BIKE, Back-Flip and Black-Gray, and propose a new decoder, called Black-Gray-Flip, that achieves the same DFR as the two previously used decoders while being almost twice as fast. Finally, we propose an algorithm for inversion of binary polynomials in a polynomial ring used in BIKE-2, the second variant of BIKE. Our implementation of the inversion significantly outperforms previously used algorithms. With this and the fact that the bandwidth requirement for BIKE-2 is the smallest among the three variants, BIKE-2 is positioned as the preferable variant of BIKE. The second part of this thesis studies the Legendre pseudorandom function (PRF) which is proposed to be used in the context of blockchains. We present a new algorithm for cryptanalysis of the Legendre PRF. The complexity of our algorithm is lower than the previous best known algorithm. Moreover, we show the results of breaking three Legendre PRF challenges posed by the Ethereum foundation. The most difficult challenge that we solved set the new record which is not broken so far.

EPFL2020

Revisiting Iterated Attacks in the Context of Decorrelation Theory

Asli Bay, Atefeh Mashatan, Serge Vaudenay

Iterated attacks are comprised of iterating adversaries who can make d plaintext queries, in each iteration to compute a bit, and are trying to distinguish between a random cipher C and the perfect cipher C* based on all bits. Vaudenay showed that a 2d-decorrelated cipher resists to iterated attacks of order d. when iterations have almost no common queries. Then, he first asked what the necessary conditions are for a cipher to resist a non-adaptive iterated attack of order d. I.e., whether decorrelation of order 2d-1 could be sufficient. Secondly, he speculated that repeating a plaintext query in different iterations does not provide any advantage to a non-adaptive distinguisher. We close here these two long-standing open problems negatively. For those questions, we provide two counter-intuitive examples. We also deal with adaptive iterated adversaries who can make both plaintext and ciphertext queries in which the future queries are dependent on the past queries. We show that decorrelation of order 2d protects against these attacks of order d. We also study the generalization of these distinguishers for iterations making non-binary outcomes. Finally, we measure the resistance against two well-known statistical distinguishers, namely, differential-linear and boomerang distinguishers and show that 4-decorrelation degree protects against these attacks.

2014

Analysis of the BIKE post-quantum cryptographic protocols and the Legendre pseudorandom function

Dusan Kostic

EPFL2020