By associating Fermat's Little Theorem based $GF(2^n)$ inversion algorithms with the multiplicative Norm function, we present an additive Trace based $GF(2^n)$ inversion algorithm. For elements with Trace value 0, it needs 1 less multiplication operation than Fermat's Little Theorem based algorithms in some $GF(2^n)$s.

Due to growing commercial applications like Blockchain, the performance of large-integer arithmetic is the focus of both academic and industrial research. IBM introduced a new integer fused multiply-add instruction in z14, called VMSL, to accelerate such workloads. Unlike their floating-point counterparts, there are a variety of integer fused multiply-add instruction designs. VMSL multiplies two pairs of radix $2^{56}$ inputs, sums the two results together with an additional 128-bit input, and stores the resulting 128-bit value in a vector register. In this paper, we will describe the unique features of VMSL, the ways in which it is inherently more efficient than alternative specifications, in particular by enabling multiple carry strategies. We will then look at the issues we encountered implementing Montgomery Modular Multiplication for Elliptic Curve Cryptography on z14, including radix choice, mixed radices, instruction selection to trade instruction count for latency, and VMSL-specific optimizations for Montgomery-friendly moduli. The best choices resulted in a 20% increase in throughput.

This study is an attempt in quest of the fastest hardware algorithms for the computation of the evaluation component of verifiable delay functions (VDFs), $a^{2^T} \bmod N$, proposed for use in various distributed protocols, in which no party is assumed to compute it significantly faster than other participants. To this end, we propose a class of modular squaring algorithms suitable for low-latency ASIC implementations. The proposed algorithms aim to achieve highest levels of parallelization that have not been explored in previous works in the literature, which usually pursue more balanced optimization of speed and area. For this, we utilize redundant representations of integers and introduce three modular squaring algorithms that work with integers in redundant forms: i) Montgomery algorithm, ii) memory-based algorithm and iii) direct reduction algorithm for fixed moduli. All algorithms enable $O(\log k)$ depth circuit implementations, where $k$ is the bit-size of the modulus $N$ in the VDF function. We analyze and compare gate level-circuits of the proposed algorithms and provide estimates for their critical path delay and gate count.

In this work, we introduce the notion of puncturable witness pseudorandom function (pWPRF) which is a stronger variant of WPRF proposed by Zhandry, TCC 2016. The punctured technique is similar to what we have seen for puncturable PRFs and is capable of extending the applications of WPRF. Specifically, we construct a semi-adaptively secure offline witness encryption (OWE) scheme using a pWPRF, an indistinguishability obfuscation (iO) and a symmetric-key encryption (SKE), which enables us to encrypt messages along with NP statements. We show that replacing iO with extractability obfuscation, the OWE turns out to be an extractable offline witness encryption scheme. To gain finer control over data, we further demonstrate how to convert our OWEs into offline functional witness encryption (OFWE) and extractable OFWE. All of our OWEs and OFWEs produce an optimal size ciphertext, in particular, encryption of a message is as small as the size of the message plus the security parameter multiplied with a constant, which is optimal for any public-key encryption scheme. On the other hand, in any previous OWE, the size of a ciphertext increases polynomially with the size of messages. Finally, we show that the WPRF of Pal et al. (ACISP 2019) can be extended to a pWPRF and an extractable pWPRF.

In a recent work, Kumar, Meka, and Sahai (FOCS 2019) introduced the notion of bounded collusion protocols (BCPs), in which $N$ parties wish to compute some joint function $f:(\{0,1\}^n)^N\to\{0,1\}$ using a public blackboard, but such that only $p$ parties may collude at a time. This generalizes well studied models in multiparty communication complexity, such as the number-in-hand (NIH) and number-on-forehead (NOF) models, which are just endpoints on this rich spectrum. We construct explicit hard functions against this spectrum, and achieve a tradeoff between collusion and complexity. Using this, we obtain improved leakage-resilient secret sharing schemes against bounded collusion protocols. Our main tool in obtaining hard functions against BCPs are explicit constructions of leakage resilient extractors against BCPs for a wide range of parameters. Kumar et al. (FOCS 2019) studied such extractors and called them cylinder intersection extractors. In fact, such extractors directly yield correlation bounds against BCPs. We focus on the following setting: the input to the extractor consists of $N$ independent sources of length $n$, and the leakage function Leak $:(\{0,1\}^n)^N\to\{0,1\}^\mu\in\mathcal{F}$ is a BCP with some collusion bound $p$ and leakage (output length) $\mu$. While our extractor constructions are very general, we highlight some interesting parameter settings: 1. In the case when the input sources are uniform, and $p=0.99N$ parties collude, our extractor can handle $n^{\Omega(1)}$ bits of leakage, regardless of the dependence between $N,n$. The best NOF lower bound (i.e., $p=N-1$) on the other hand requires $N<\log n$ even to handle $1$ bit of leakage. 2. Next, we show that for the same setting as above, we can drop the entropy requirement to $k=$ polylog $n$, while still handling polynomial leakage for $p=0.99N$. This resolves an open question about cylinder intersection extractors raised by Kumar et al. (FOCS 2019), and we find an application of such low entropy extractors in a new type of secret sharing. We also provide an explicit compiler that transforms any function with high NOF (distributional) communication complexity into a leakage-resilient extractor that can handle polylogarithmic entropy and substantially more leakage against BCPs. Thus any improvement of NOF lower bounds will immediately yield better leakage-resilient extractors. Using our extractors against BCPs, we obtain improved $N$-out-of-$N$ leakage-resilient secret sharing schemes. The previous best scheme from Kumar et al. (FOCS 2019) required share size to grow exponentially in the collusion bound, and thus cannot efficiently handle $p=\omega(\log N)$. Our schemes have no dependence of this form, and can thus handle collusion size $p=0.99N$.

In this work we first provide a framework for defining a large subset of pairing-based digital signature schemes which we call Partially Structure-Preserving Signature (PSPS) schemes. PSPS schemes are similar in nature to structure-preserving signatures with the exception that in these schemes messages are scalars from $\Z^n_p$ instead of being source group elements. This class encompasses various existing schemes which have a number of desirable features which makes them an ideal building block for many privacy-preserving cryptographic protocols. They include the widely-used schemes of Camenisch-Lysyanskaya (CRYPTO 2004) and Pointcheval-Sanders (CT-RSA 2016). We then provide various impossibility and lower bound results for variants of this class. Our results include bounds for the signature and verification key sizes as well as lower bounds for achieving strong unforgeability. We also give a generic framework for transforming variants of PSPS schemes into structure-preserving ones. As part of our contribution, we also give a number of optimal PSPS schemes which may be of independent interest. Our results aid in understanding the efficiency of pairing-based signature schemes and show a connection between this class of signature schemes and structure-preserving ones.

Decentralized and permissionless ledgers offer an inherently low transaction rate, as a result of their consensus protocol demanding the storage of each transaction on-chain. A prominent proposal to tackle this scalability issue is to utilize off-chain protocols, where parties only need to post a limited number of transactions on-chain. Existing solutions can roughly be categorized into: (i) application-specific channels (e.g., payment channels), offering strictly weaker functionality than the underlying blockchain; and (ii) state channels, supporting arbitrary smart contracts at the cost of being compatible only with the few blockchains having Turing-complete scripting languages (e.g., Ethereum). In this work, we introduce and formalize the notion of generalized channels allowing users to perform any operation supported by the underlying blockchain in an off-chain manner. Generalized channels thus extend the functionality of payment channels and relax the definition of state channels. We present a concrete construction compatible with any blockchain supporting transaction authorization, time-locks and constant number of Boolean $\land$ and $\lor$ operations -- requirements fulfilled by many (non-Turing-complete) blockchains including the popular Bitcoin. To this end, we leverage adaptor signatures -- a cryptographic primitive already used in the cryptocurrency literature but formalized as a standalone primitive in this work for the first time. We formally prove the security of our generalized channel construction in the Universal Composability framework. As an important practical contribution, our generalized channel construction outperforms the state-of-the-art payment channel construction, the Lightning Network, in efficiency. Concretely, it halves the off-chain communication complexity and reduces the on-chain footprint in case of disputes from linear to constant in the number of off-chain applications funded by the channel. Finally, we evaluate the practicality of our construction via a prototype implementation and discuss various applications including financially secured fair two-party computation.

This paper provides a theoretical background for a new consensus model called Proof of Review (PoR), which extends Algorand’s blockchain consensus model and reproduces the human mechanism for analyzing reviews through analysis and reputation. Our protocol derives the trustworthiness of a participant’s reputation through a consensus of these reviews. In this new protocol, we combined concepts from proof of stake and proof of reputation to ensure a blockchain system comes to consensus on an honest (non-malicious) congruent review. Additionally, we formally prove this protocol provides further security by using a reputation-based stake instead of token-based, using theorems and proofs. We introduce new concepts in using reputation as a stake, where reputation must be earned and can never be purchased, spent, or traded. The stake is calculated based on the reputation -- not tokens -- of a user in the system proportional to the total reputation in the system at the beginning of each round. A round is a set of steps that conclude in a block being added to the blockchain. We also introduce blacklisting as a decisive action where, after a vote, an honest user will blacklist a leader (the participant elected with their proposed block) or reviewer for some egregious maliciousness, preventing further participation. Five steps are defined that overlay Algorand’s steps: (1) Evaluation of reviews; (2) Selection of the round’s leader and their associated block containing reviews; (3) Re-evaluation of reviews; (4) Block of reviews agreement (block decision); (5) Block addition to the ledger and reputation adjustment. At every step, verifiers are selected randomly for evaluating the reviews and are anonymous to each other. We provided properties related to reputation and formally proved (Honest Leader, Blacklisting Liveliness and Correctness, Evaluation of the Evaluator’s Confidence, Gradual Gain, and Swift Loss of Reputation). Finally, we discuss several types of attacks that are common for Proof of Stake and Reputation systems, and how our Proof of Review model (in extending Algorand) addresses those attacks. Furthermore, PoR mitigates several kinds of fake reviews (e.g. spam, trolling, etc.) through analysis. Preliminary experimental results show that payment transactions have similar completion times, regardless of the stake being tokens or reputation, the mechanism for adjusting reputation reflects expected behavior, and the accuracy of the review evaluation (using sentimental analysis) is substantiated for the dataset given. With this new model, we let the technology evaluate a review and secure it on a blockchain using a reputation-based stake. Proof of Review also provides third party access to reviewer’s reputation along with their respective evaluated reviews, a functionality that could benefit numerous industries including retail, academia, hospitality, corporate, and more. Applications could expedite and validate many different types of processes that rely on a form of review and the analysis of that review.

Zk-SNARKs, as the most efficient NIZK arguments in terms of proof size and verification, are ubiquitously deployed in practice. In applications like Hawk [S&P'16], Gyges [CCS'16], Ouroboros Crypsinous [S&P'19], the underlying zk-SNARK is lifted to achieve Black-Box Simulation Extractability (BB-SE) under a trusted setup phase. To mitigate the trust in such systems, we propose $\texttt{Tiramisu}$, as a construction to build NIZK arguments that can achieve $\textit{updatable BB-SE}$, which we define as a new variant of BB-SE. This new variant allows $\textit{updating}$ the public parameters, therefore eliminating the need for a trusted third party, while unavoidably relies on a $\textit{non-black-box}$ extraction algorithm in the setup phase. In the cost of one-time individual CRS update by the parties, this gets around a known impossibility result by Bellare et al. from ASIACRYPT'16, which shows that BB extractability cannot be achieved with subversion ZK (ZK without trusting a third party). $\texttt{Tiramisu}$ uses an efficient public-key encryption with updatable keys which may be of independent interest. We instantiate $\texttt{Tiramisu}$, implement the overhead, and present efficient BB-SE zk-SNARKs with updatable parameters that can be used in various applications while allowing the end-users to update the parameters and eliminate the needed trust.

In this work we study bounded collusion protocols (BCPs) recently introduced in the context of secret sharing by Kumar, Meka, and Sahai (FOCS 2019). These are multi-party communication protocols on $n$ parties where in each round a subset of $p$-parties (the collusion bound) collude together and write a function of their inputs on a public blackboard. BCPs interpolate elegantly between the well-studied number-in-hand (NIH) model ($p=1$) and the number-on-forehead (NOF) model ($p=n-1$). Motivated by questions in communication complexity, secret sharing, and pseudorandomness we investigate BCPs more thoroughly, answering several questions about them. * We prove a polynomial (in the input-length) lower bound for an explicit function against BCPs where any constant fraction of players can collude. Previously, nontrivial lower bounds were known only when the collusion bound was at most logarithmic in the input-length (owing to bottlenecks in NOF lower bounds). * For all $t \leq n$, we construct efficient $t$-out-of-$n$ secret sharing schemes where the secret remains hidden even given the transcript of a BCP with collusion bound $O(t/\log t)$. Prior work could only handle collusions of size $O(\log n)$. Along the way, we construct leakage-resilient schemes against disjoint and adaptive leakage, resolving a question asked by Goyal and Kumar (STOC 2018). * An explicit $n$-source cylinder intersection extractor whose output is close to uniform even when given the transcript of a BCP with a constant fraction of parties colluding. The min-entropy rate we require is $0.3$ (independent of collusion bound $p \ll n$). Our results rely on a new class of exponential sums that interpolate between the ones considered in additive combinatorics by Bourgain (Geometric and Functional Analysis 2009) and Petridis and Shparlinski (Journal d'Analyse Mathématique 2019).

Directed Acyclic Graph (DAG) is becoming an intriguing direction for distributed ledger structure due to its great potential in improving the scalability of distributed ledger systems. Among existing DAG-based ledgers, one promising category is transaction DAG, namely, treating each transaction as a graph vertex. In this paper, we propose Haootia, a novel two-layer framework of consensus, with a ledger in the form of a transaction DAG built on top of a delicately designed PoW-based backbone chain. By elaborately devising the principle of transaction linearizations, we achieve a secure and scalable DAG-based consensus. By implementing Haootia, we conclude that, with a rotating committee of size 46 and a confirmation latency around 20 seconds, Haootia achieves a throughput around 7500 TPS which is overwhelming compared with all formally analyzed DAG-based consensus schemes to date and all existing non-DAG-based ones to our knowledge.

In this work, we prove that Interpose PUF is learnable in the PAC model. First, we show that Interpose PUF can be approximated by a Linear Threshold Function~(LTF), assuming the interpose bit to be random. We translate the randomness in the interpose bit to classification noise of the hypothesis. Using classification noise model, we prove that the resultant LTF can be learned with number of labelled examples~(challenge response pairs) polynomial in the number of stages and PAC model parameters.

Stateful hash-based signature schemes are among the most efficient approaches for post-quantum signature schemes. Although not suitable for general use, they may be suitable for some use cases on constrained devices. LMS and XMSS are hash-based signature schemes that are conjectured to be quantum secure. In this work, we compared multiple instantiations of both schemes on an ARM Cortex-M4. More precisely, we compared performance, stack consumption, and other figures for key generation, signing and verifying. To achieve this, we evaluated LMS and XMSS using optimised implementations of SHA-256, SHAKE256, Gimli-Hash, and different variants of Keccak. Furthermore, we present slightly optimised implementations of XMSS achieving speedups of up to 3.11x for key generation, 3.11x for signing, and 4.32x for verifying.

Secure enclave architectures have become prevalent in modern CPUs and enclaves provide a flexible way to implement various hardware-assisted security services. But special-purpose security chips can still have advantages. Interestingly, dedicated security chips can also assist enclaves and improve their security.

Distributed Auditing Proofs of Liabilities (DAPOL) provides a novel zero knowledge proof solution to a particular class of auditing cases, in which we assume that the audited entity does not have any incentive to increase its liabilities or obligations. There are numerous domains requiring such an auditing feature, including proving financial solvency, transparent fundraising campaigns and accurate lottery jackpot amounts. Additionally, the algorithm provides a solution to official reports, such as in COVID-19 published daily cases, unemployment rate announcements and decentralized product/service rating reviews. Interestingly, it can also be used as a cryptographic primitive for novel e-voting systems (i.e., disapproval voting and counting dislikes), and for innovative private syndicated loan/insurance solutions, new methods for decentralized credit scoring and user ranking, among the others. Compared to conventional auditor-based approaches, DAPOL provides a privacy preserving mechanism for users to validate their vote or amount inclusion in the reported total of liabilities/obligations and complements the traditional validation performed by the auditors by adding extra privacy and fairness guarantees. The recommended approach combines previously known cryptographic techniques to provide a layered solution with predefined levels of privacy in the form of gadgets. The backbone of this proposal is based on the enhanced Maxwell Merkle-tree construction and is extended using zero knowledge proofs, sparse trees, balance splitting tricks, efficient padding, verifiable random functions, deterministic key derivation functions and the range proof techniques from Provisions and ZeroLedge solvency protocols, respectively.

The Meet-in-the-Middle (MITM) preimage attack is highly effective in breaking the preimage resistance of many hash functions, including but not limited to the full MD5, HAVAL, and Tiger, and reduced SHA-0/1/2. It was also shown to be a threat to hash functions built on block ciphers like AES by Sasaki in 2011. Recently, such attacks on AES hashing modes evolved from merely using the freedom of choosing the internal state to also exploiting the freedom of choosing the message state. However, detecting such attacks especially those evolved variants is difficult. In previous works, the search space of the configurations of such attacks is limited, such that manual analysis is practical, which results in sub-optimal solutions. In this paper, we remove artificial limitations in previous works, formulate the essential ideas of the construction of the attack in well-defined ways, and translate the problem of searching for the best attacks into optimization problems under constraints in Mixed-Integer-Linear-Programming (MILP) models. The MILP models capture a large solution space of valid attacks; and the objectives of the MILP models are attack configurations with the minimized computational complexity. With such MILP models and using the off-the-shelf solver, it is efficient to search for the best attacks exhaustively. As a result, we obtain the first attacks against the full (5-round) and an extended (5.5-round) version of Haraka-512 v2, and 8-round AES-128 hashing modes, as well as improved attacks covering more rounds of Haraka-256 v2 and other members of AES and Rijndael hashing modes.

The design of software countermeasures against active and passive adversaries is a challenging problem that has been addressed by many authors in recent years. The proposed solutions adopt a theoretical foundation (such as a leakage model) but often do not offer concrete reference implementations to validate the foundation. Contributing to the experimental dimension of this body of work, we propose a customized processor called SKIVA that supports experiments with the design of countermeasures against a broad range of implementation attacks. Based on bitslice programming and recent advances in the literature, SKIVA offers a flexible and modular combination of countermeasures against power-based and timing-based side-channel leakage and fault injection. Multiple configurations of side-channel protection and fault protection enable the programmer to select the desired number of shares and the desired redundancy level for each slice. Recurring and security-sensitive operations are supported in hardware through custom instruction-set extensions. The new instructions support bitslicing, secret-share generation, redundant logic computation, and fault detection. We demonstrate and analyze multiple versions of AES from a side-channel analysis and a fault-injection perspective, in addition to providing a detailed performance evaluation of the protected designs. To our knowledge, this is the first validated end-to-end implementation of a modular bitslice-oriented countermeasure.

An important selling point for the RISC-V instruction set is the separation between ISA and the implementation of the ISA, leading to flexibility in the design. We argue that for secure implementations, this flexibility is often a vulnerability. With a hardware attacker, the side-effects of instruction execution cannot be ignored. As a result, a strict separation between the ISA interface and implementation is undesirable. We suggest that secure ISA may require additional implementation constraints. As an example, we describe an instruction-set for the development of power side-channel resistant software.

Blockchains are gaining traction and acceptance, not just for cryptocurrencies, but increasingly as an architecture for distributed computing. In this work we seek solutions that allow a \emph{public} blockchain to act as a trusted long-term repository of secret information: Our goal is to deposit a secret with the blockchain, specify how it is to be used (e.g., the conditions under which it is released), and have the blockchain keep the secret and use it only in the specified manner (e.g., release only it once the conditions are met). This simple functionality enables many powerful applications, including signing statements on behalf of the blockchain, using it as the control plane for a storage system, performing decentralized program-obfuscation-as-a-service, and many more. Using proactive secret sharing techniques, we present a scalable solution for implementing this functionality on a public blockchain, in the presence of a mobile adversary controlling a small minority of the participants. The main challenge is that, on the one hand, scalability requires that we use small committees to represent the entire system, but, on the other hand, a mobile adversary may be able to corrupt the entire committee if it is small. For this reason, existing proactive secret sharing solutions are either non-scalable or insecure in our setting. We approach this challenge via "player replaceability", which ensures the committee is anonymous until after it performs its actions. Our main technical contribution is a system that allows sharing and re-sharing of secrets among the members of small dynamic committees, without knowing who they are until after they perform their actions and erase their secrets. Our solution handles a fully mobile adversary corrupting roughly 1/4 of the participants at any time, and is scalable in terms of both the number of parties and the number of time intervals.

The advances in machine learning have revealed its great potential for emerging mobile applications such as face recognition and voice assistant. Models trained via a Neural Network (NN) can offer accurate and efficient inference services for mobile users. Unfortunately, the current deployment of such service encounters privacy concerns. Directly offloading the model to the mobile device violates model privacy of the model owner, while feeding user input to the service compromises user privacy. To address this issue, we propose, tailor, and evaluate Leia, a lightweight cryptographic NN inference system at the edge. Unlike prior cryptographic NN inference systems, Leia is designed with two mobile-friendly perspectives. First, Leia leverages the paradigm of edge computing wherein the inference procedure keeps the model closer to the mobile user to foster low latency service. Specifically, Leia's architecture consists of two non-colluding edge services to obliviously perform NN inference on the encoded user data and model. Second, Leia's realization makes the judicious use of potentially constrained computational and communication resources in edge devices. In particular, Leia adapts the Binarized Neural Network (BNN), a trending flavor of NN model with low memory footprint and computational cost, and purely chooses the lightweight secret sharing techniques to develop secure blocks of BNN. Empirical validation executed on Raspberry Pi confirms the practicality of Leia, showing that Leia can produce a prediction result with 97% accuracy by 4 seconds in the edge environment.

Motivated by a fundamental paradigm in cryptography, we consider a recent variant of the classic problem of bounding the distinguishing advantage between a random function and a random permutation. Specifically, we consider the problem of deciding whether a sequence of $q$ values was sampled uniformly with or without replacement from $[N]$, where the decision is made by a streaming algorithm restricted to using at most s bits of internal memory. In this work, the distinguishing advantage of such an algorithm is measured by the KL divergence between the distributions of its output as induced under the two cases. We show that for any $ s=\Omega( \log N) $ the distinguishing advantage is upper bounded by $ O(q\cdot s/N) $, and even by $ O(q\cdot s/N \log N) $ when $ q\le N^{1-\epsilon} $ for any constant $ \epsilon>0 $ where it is nearly tight with respect to the KL divergence.

At Crypto ’99, Nguyen and Stern described a lattice based algorithm for solving the hidden subset sum problem, a variant of the classical subset sum problem where the n weights are also hidden. While the Nguyen-Stern algorithm works quite well in practice for moderate values of n, we argue that its complexity is actually exponential in n; namely in the final step one must recover a very short basis of a n-dimensional lattice, which takes exponential-time in n, as one must apply BKZ reduction with increasingly large block-sizes. In this paper, we describe a variant of the Nguyen-Stern algorithm that works in polynomial-time. The first step is the same orthogonal lattice attack with LLL as in the original algorithm. In the second step, instead of applying BKZ, we use a multivariate technique that recovers the short lattice vectors and finally the hidden secrets in polynomial time. Our algorithm works quite well in practice, as we can reach n=250 in a few hours on a single PC.

Verifying security of advanced cryptographic primitives such as attribute-based encryption (ABE) is often difficult. In this work, we show how to break eleven schemes: two single-authority and nine multi-authority (MA) ABE schemes. Notably, we break DAC-MACS, a highly-cited multi-authority scheme, published at TIFS. This suggests that, indeed, verifying security of complex schemes is complicated, and may require simpler tools. The multi-authority attacks also illustrate that mistakes are made in transforming single-authority schemes into multi-authority ones. To simplify verifying security, we systematize our methods to a linear approach to analyzing generic security of ABE. Our approach is not only useful in analyzing existing schemes, but can also be applied during the design and reviewing of new schemes. As such, it can prevent the employment of insecure (MA-)ABE schemes in the future.

Fault Template Analysis (FTA) has been shown as a powerful tool for attacking cryptosystems and exposing vulnerabilities which were previously not reported in existing literature. Fault templates can be utilized for attacking block ciphers in middle rounds which were known prior to be resistant against fault attacks. In this paper we revisit the potent of fault templates and show a more systematic methodology to develop fault templates of Boolean circuits using a well known concept in design verification, namely positive Davio's decomposition. We show that the improved FTAs, called FTA2.0, can be used to fault analyze block ciphers in the middle rounds using as few as two bit-flip faults. Further, it can be used to attack TI-implemented block ciphers by considering a Double Bit Upset (DBU) fault in a target share bit. The attack shows that varying the latency of the fault the adversary can obtain unmasked bits and can recover the secret key.

Synchronous consensus protocols, by definition, have a worst-case commit latency that depends on the bounded network delay. The notion of optimistic responsiveness was recently introduced to allow synchronous protocols to commit instantaneously when some optimistic conditions are met. In this work, we revisit this notion of optimistic responsiveness and present optimal latency results. We present a lower bound for Byzantine Broadcast that relates the latencies of optimistic and synchronous commits when the designated sender is honest and while the optimistic commit can tolerate some faults. We then present two matching upper bounds for tolerating f faults out of n = 2f +1 parties. Our first upper bound result achieves optimal optimistic and synchronous commit latencies when the designated sender is honest and the optimistic commit can tolerate some faults. Our second upper bound result achieves optimal optimistic and synchronous commit latencies when the designated sender is honest but the optimistic commit does not tolerate any faults. The presence of matching lower and upper bound results make both of the results tight for n = 2f + 1. Our upper bound results are presented in a state machine replication setting with a steady state leader who is replaced with a view-change protocol when they do not make progress. For this setting, we also present an optimistically responsive protocol where the view-change protocol is optimistically responsive too.

Sharing a documents with a business partner is not always easy. since the sender often need to send sensitive information. and he want to ensure the integrity and the secrecy of the document. And in the same time. he wants to insure that only the specific individual or the recipients are the only one who can view it. So people tend to use some encryption software. or protecting the document with some sort of password. and then share the password with the recipient to make sure he is the only one who can view the document. But Unfortunately in many situations this method will not work. for a particular reason. and that is once the sender send an email. the email will start his journey into the company's network. and it will pass through many appliances. such Firewalls, Exchange servers and most likely Sandboxes. And there is one feature in sandboxes that we are interested in. once the sandbox sees an encrypted file or a protected file. it will immediately stop the email and quarantine it. because the sandbox couldn’t scan it. or couldn’t ensure if it’s malicious or not. so it will stop it for further analysis or a manual analysis depending on the procedures there. And such an action could stop a valid business transaction. and it could cause some business interruption. In this paper we will introduce a scheme for allowing the share of protected files. and analyzing them through Sandboxes. and in the same time no one can view it except for the authorized people.

Payment Channel Networks (PCNs) have been a promising approach to scale blockchains. However, PCNs have limited liquidity: large-amount or multi-hop payments may fail. The major threat of PCNs liquidity is payment griefing, where the adversary who acts as the payee keeps withholding the payment, so that coins involved in the payment cannot be used for routing other payments before the payment expires. Payment griefing gives adversaries a chance to launch the congestion attack, where the adversary griefs a large number of payments and paralyses the entire PCN. Understanding congestion attacks, including their strategies and impact, is crucial for designing PCNs with better liquidity guarantees. However, existing research has only focused on the specific attacking strategies and specific aspects of their impact on PCNs. We fill this gap by studying the general congestion attack. Compared to existing attack strategies, in our framework each step serves an orthogonal purpose and is customisable, allowing the adversary to focus on different aspects of the liquidity. To evaluate the attack's impact, we propose a generic method of quantifying PCNs' liquidity and effectiveness of the congestion attacks. We evaluate our general congestion attacks on Bitcoin's Lightning Network, and show that with direct channels to 1.5% richest nodes, and $\sim$ 0.0096 BTC of cost, the adversary can launch a congestion attack that locks 47% ($\sim$280 BTC) coins in the network; reduces success rate of payments by 16.0%$\sim$60.0%; increases fee of payments by 4.5%$\sim$16.0%; increases average attempts of payments by 42.0%$\sim$115.3%; and increase the number of bankruptcy nodes (i.e., nodes with insufficient balance for making normal-size payments) by 26.6%$\sim$109.4%, where the amounts of payments range from 0.001 to 0.019 BTC.

We report on the concrete cryptanalysis of LEDAcrypt, a 2nd Round candidate in NIST's Post-Quantum Cryptography standardization process and one of 17 encryption schemes that remain as candidates for near-term standardization. LEDAcrypt consists of a public-key encryption scheme built from the McEliece paradigm and a key-encapsulation mechanism (KEM) built from the Niederreiter paradigm, both using a quasi-cyclic low-density parity-check (QC-LDPC) code. In this work, we identify a large class of extremely weak keys and provide an algorithm to recover them. For example, we demonstrate how to recover 1 in $2^{47.72}$ of LEDAcrypt's keys using only $2^{18.72}$ guesses at the 256-bit security level. This is a major, practical break of LEDAcrypt. Further, we demonstrate a continuum of progressively less weak keys (from extremely weak keys up to all keys) that can be recovered in substantially less work than previously known. This demonstrates that the imperfection of LEDAcrypt is fundamental to the system's design.

We present an optimization of Lagrange's algorithm for lattice basis reduction in dimension 2. The optimized algorithm is proven to be correct and to always terminate with quadratic complexity; it uses more iterations on average than Lagrange's algorithm, but each iteration is much simpler to implement, and faster. The achieved speed is such that it makes application of the speed-up on ECDSA and EC Schnorr signatures described by Antipa et al worthwhile, even for very fast curves such as Ed25519. We applied this technique to signature verification in Curve9767, and reduced verification time by 30 to 33% on both small (ARM Cortex M0+ and M4) and large (Intel Coffee Lake with AVX2) architectures.

We design a consecution of protocols which allows organizations to have secure strong access control of their users to their desktop machines based on biometry. It provides both strong secure authentication and privacy. Moreover, our mechanism allows the system admins to grant a various level of access to their end-users by fine tuning access control policy. Our system implements privacy-by-design. It separates biometric data from identity information. It is practical: we fully implemented our protocols as a proof of concept for a hospital. We use a 3D fingervein scanner to capture the biometric data of the user on a Raspberry Pi. For the biometry part, we developed an optimal way to aggregate scores using sequential distinguishers. It trades desired FAR and FRR against an average number of biometric captures.

In a quantum money scheme, a bank can issue money that users cannot counterfeit. Similar to bills of paper money, most quantum money schemes assign a unique serial number to each money state, thus potentially compromising the privacy of the users of quantum money. However in a quantum coins scheme, just like the traditional currency coin scheme, all the money states are exact copies of each other, providing a better level of privacy for the users. A quantum money scheme can be private, i.e., only the bank can verify the money states, or public, meaning anyone can verify. In this work, we propose a way to lift any private quantum coin scheme -- which is known to exist based on the existence of one-way functions, due to Ji, Liu, and Song (CRYPTO'18) -- to a scheme that closely resembles a public quantum coin scheme. Verification of a new coin is done by comparing it to the coins the user already possesses, by using a projector on to the symmetric subspace. No public coin scheme was known prior to this work. It is also the first construction that is very close to a public quantum money scheme and is provably secure based on standard assumptions. The lifting technique when instantiated with the private quantum coins scheme, due to Mosca and Stebila 2010, gives rise to the first construction that is very close to an inefficient unconditionally secure public quantum money scheme.

Computing on data in a manner that preserve the privacy is of growing importance. Multi-Party Computation (MPC) and Homomorphic Encryption (HE) are two cryptographic techniques for privacy-preserving computations. In this work, we have developed efficient UC-secure multiparty protocols for matrix multiplications and two-dimensional convolutions. We built upon the SPDZ framework and integrated the state-of-the-art HE algorithms for matrix multiplication. Our protocol achieved communication cost linear only in the input and output dimensions and not on the number of multiplication operations. We eliminate the ``triple sacrifice'' step of SPDZ to improve efficiency and simplify the zero-knowledge proofs. We implemented our protocols and benchmarked them against the SPDZ LowGear variant (Keller et al. Eurocrypt'18). For multiplying two square matrices of size 128, we reduced the communication cost from 1.54 GB to 12.46 MB, an improvement of over two orders of magnitude that only improves with larger matrix sizes. For evaluating all convolution layers of the ResNet-50 neural network, the communication reduces cost from 5 TB to 41 GB.

Pointcheval-Sanders (PS) signatures are well-studied in the literature and have found use within e.g. threshold credential schemes and redactable anonymous credential schemes. The present work leverages a mapping between PS signatures and a related class of polynomial-based signatures to construct multiple new signature/credential schemes. Specifically, new protocols for multi-message signatures, sequential aggregate signatures, signatures for message commitments, redactable signatures, and unlinkable redactable signatures are presented. A redactable anonymous credential scheme is also constructed. All original protocols employ constant-sized secret keys rather than linear-sized (in the number of messages/attributes). Security properties of the new protocols are analysed and a general discussion of security properties for both PS signatures and the new schemes is provided.

Variant secret sharing schemes deriving from Shamir's threshold secret sharing protocol are presented. Results include multi-secret sharing protocols using shares with $O(1)$ elements, independent of the number of secrets. The new schemes achieve a weaker notion of security (they're secure rather than strongly secure) but maintain a property called $K$-privacy (inspired by $k$-anonymity). $K$-privacy ensures that all secrets remain private with respect to a subset of the secret space, though the particular subset providing privacy may vary among adversaries that acquire distinct sub-threshold sets of shares. Depending on the number of secrets and the protocol details, secure $K$-private multi-secret sharing schemes may be ``almost'' strongly secure or may remain merely secure and $K$-private - a difference captured by the notion of $K$-security. Novel applications of the multi-secret sharing schemes are presented, realising a primitive called a switched threshold signature. Switched threshold signatures have the quirky property that aggregating a threshold number of signatures of one type (e.g. Pointcheval-Sanders signatures) ``switches'' the signatures into a master signature of a different type. Collectively these results may permit efficiencies within, e.g., threshold credential issuance protocols.

The information ratio of an access structure is an important parameter for quantifying the efficiency of the best secret sharing scheme (SSS) realizing it. The most common security notion is perfect security. The following relaxations, in increasing level of security, have been presented in the literature: quasi-perfect, almost-perfect and statistical. Understanding the power of relaxing the correctness and privacy requirements in the efficiency of SSSs is a long-standing open problem. In this article, we introduce and study an extremely relaxed security notion, called partial security, for which it is only required that any qualified set gains strictly more information about the secret than any unqualified one. We refer to this gap as the nominal capacity. We quantify the efficiency of such schemes using a parameter called partial information ratio. It is defined to be the same as the (standard) information ratio, except that we divide the largest share entropy by nominal capacity instead of the secret entropy. Despite this modification, partial security turns out weaker than the weakest mentioned non-perfect security notion, i.e., quasi-perfect security. We present three main results in this paper. First, we prove that partial and perfect information ratios coincide for the class of linear SSSs. Consequently, for this class, information ratio is invariant with respect to all security notions. Second, by viewing a partial SSS as a wiretap channel, we prove that for the general (i.e., non-linear) class of SSSs, partial and statistical information ratios are equal. Consequently, for this class, information ratio is invariant with respect to all non-perfect security notions. Third, we show that partial and almost-perfect information ratios do not coincide for the class of mixed-linear schemes (i.e., schemes constructed by combining linear schemes with different underlying finite fields). Our first result strengthens the previous decomposition theorems for constructing perfect linear schemes. Our second result leads to a very strong decomposition theorem for constructing general (i.e., non-linear) statistical schemes. Our third result provides a rare example of the effect of imperfection on the efficiency of SSSs for a certain class of schemes.

New cryptographic techniques such as homomorphic encryption (HE) allow computations to be outsourced to and evaluated blindfolded in a resourceful cloud. These computations often require private data owned by multiple participants, engaging in joint evaluation of some functions. For example, Genome-Wide Association Study (GWAS) is becoming feasible because of recent proliferation of genome sequencing technology. Due to the sensitivity of genomic data, these data should be encrypted using different keys. However, supporting computation on ciphertexts encrypted under multiple keys is a non-trivial task. In this paper, we present a comprehensive survey on different state-of-the-art cryptographic techniques and schemes that are commonly used. We review techniques and schemes including Attribute-Based Encryption (ABE), Proxy Re-Encryption (PRE), Threshold Homomorphic Encryption (ThHE), and Multi-Key Homomorphic Encryption (MKHE). We analyze them based on different system and security models and examine their complexities. We share lessons learned and draw observations for designing better schemes with reduced overheads.

Empowering electronic devices to support Post-Quantum Cryptography (PQC) is a challenging task. PQC introduces new mathematical elements and operations which are usually not easy to implement on standard processors. Especially for low cost and resource constraint devices, hardware acceleration is usually required. In addition, as the standardization process of PQC is still ongoing, a focus on maintaining flexibility is mandatory. To cope with such requirements, hardware/software co-design techniques have been recently used for developing complex and highly customized PQC solutions. However, while most of the previous works have developed loosely coupled PQC accelerators, the design of tightly coupled accelerators and Instruction Set Architecture (ISA) extensions for PQC have been barely explored. To this end, we present RISQ-V, an enhanced RISC-V architecture that integrates a set of powerful tightly coupled accelerators to speed up lattice-based PQC. RISQ-V efficiently reuses processor resources and reduces the amount of memory accesses. This significantly increases the performance while keeping the silicon area overhead low. We present three contributions. First, we propose a set of powerful hardware accelerators deeply integrated into the RISC-V pipeline. Second, we extended the RISC-V ISA with 29 new instructions to efficiently perform operations for lattice-based cryptography. Third, we implemented our RISQ-V in ASIC technology and on FPGA. We evaluated the performance of NewHope, Kyber, and Saber on RISQ-V. Compared to the pure software implementation on RISC-V, our co-design implementations show a speedup factor of up to 11.4 for NewHope, 9.6 for Kyber, and 2.7 for Saber. For the ASIC implementation, the energy consumption was reduced by factors of up to 9.5 for NewHope, 7.7 for Kyber, and 2.1 for Saber. The cell count of the CPU was increased by a factor of 1.6 compared to the original RISC-V design, which can be considered as a moderate increase for the achieved performance gain.

We initiate a study of pseudorandom encodings: efficiently computable and decodable encoding functions that map messages from a given distribution to a random-looking distribution. For instance, every distribution that can be perfectly and efficiently compressed admits such a pseudorandom encoding. Pseudorandom encodings are motivated by a variety of cryptographic applications, including password-authenticated key exchange, ''honey encryption'' and steganography. The main question we ask is whether every efficiently samplable distribution admits a pseudorandom encoding. Under different cryptographic assumptions, we obtain positive and negative answers for different flavors of pseudorandom encodings, and relate this question to problems in other areas of cryptography. In particular, by establishing a two-way relation between pseudorandom encoding schemes and efficient invertible sampling algorithms, we reveal a connection between adaptively secure multiparty computation for randomized functionalities and questions in the domain of steganography.

This paper revisits the Abe--Okamoto signature scheme to present a version of their signature scheme augmented with modern best practices, with major influences taken from EdDSA. Implementation guidance is given on how to reuse existing Ed25519 code.

White-box cryptography attempts to protect cryptographic secrets in pure software implementations. Due to their high utility, white-box cryptosystems (WBC) are deployed by the industry even though the security of these constructions is not well defined. A major breakthrough in generic cryptanalysis of WBC was Differential Computation Analysis (DCA), which requires minimal knowledge of the underlying white-box protection and also thwarts many obfuscation methods. To avert DCA, classic masking countermeasures originally intended to protect against highly related side-channel attacks have been proposed for use in WBC. However, due to the controlled environment of WBCs, new algebraic attacks against classic masking schemes have quickly been found. These algebraic DCA attacks break all classic masking countermeasures efficiently, as they are independent of the masking order. In this work, we propose a novel generic masking scheme that can resist both DCA and algebraic DCA attacks. The proposed scheme extends the seminal work by Ishai et al. which is probing secure and thus resists DCA, to also resist algebraic attacks. To prove the security of our scheme, we demonstrate the connection between two main security notions in white-box cryptography: probing security and prediction security. Resistance of our masking scheme to DCA is proven for an arbitrary order of protection, using the well-known strong non-interference notion by Barthe et al. Our masking scheme also resists algebraic attacks, which we show concretely for first and second-order algebraic protection. Moreover, we present an extensive performance analysis and quantify the overhead of our scheme, for a proof-of-concept protection of an AES implementation.

Fine-grained cryptography is concerned with adversaries that are only moderately more powerful than the honest parties. We will survey recent results in this relatively underdeveloped area of study and examine whether the time is ripe for further advances in it.

A division property is a generic tool to search for integral distinguishers, and automatic tools such as MILP or SAT/SMT allow us to evaluate the propagation efficiently. In the application to stream ciphers, it enables us to estimate the security of cube attacks theoretically, and it leads to the best key-recovery attacks against well-known stream ciphers. However, it was reported that some of the key-recovery attacks based on the division property degenerate to distinguishing attacks due to the inaccuracy of the division property. Three-subset division property (without unknown subset) is a promising method to solve this inaccuracy problem, and a new algorithm using automatic tools for the three-subset division property was recently proposed at Asiacrypt2019. In this paper, we first show that this state-of-the-art algorithm is not always efficient and we cannot improve the existing key-recovery attacks. Then, we focus on the feature of the three-subset division property without unknown subset and propose another new efficient algorithm using automatic tools. Our algorithm is more efficient than existing algorithms, and it can improve existing key-recovery attacks. In the application to Trivium, we show an 842-round key-recovery attack. We also show that an 855-round key-recovery attack, which was proposed at CRYPTO2018, has a critical flaw and does not work. As a result, our 842-round attack becomes the best key-recovery attack. In the application to Grain-128AEAD, we show that the known 184-round key-recovery attack degenerates to distinguishing attacks. Then, the distinguishing attacks are improved up to 189 rounds, and we also show the best key-recovery attack against 190 rounds. In the application to ACORN, we prove that the 772-round key-recovery attack at ISC2019 is in fact a constant-sum distinguisher. We then give new key-recovery attacks mounting to 773- and 774-round ACORN. We also verify the current best key-recovery attack on 892-round Kreyvium and recover the exact superpoly.

The fundamental problem in lattice-based cryptography is the hardness of the Ring-LWE, which has been based on the conjectured hardness of approximating ideal-SIVP or ideal-SVP. Though it is now widely conjectured both are hard in classical and quantum computation model” there is no sufficient attacks proposed and considered. In this paper we propose the subset quadruple attack on general structured LWE problems over any ring endowed with a positive definite inner product and an error distribution. Hence from the view of subset quadruple attacks, the error distributions of feasible non-negligible subset quadruples should be calculated to test the hardness. Sublattice pair with an ideal attack is a special case of subset quadruple attack. A lower bound for the Gaussian error distribution is proved to construct suitable feasible non-negligible sublattices. From the sublattice pair with an ideal attack we prove that the decision Poly-LWE over ${\bf Z}[x]/(x^n-p_n)$ with certain special inner products and arbitrary polynomially bounded widths of Gaussian error distributions can be solved with the polynomial time for the sufficiently large polynomially bounded modulus parameters $p_n$.\\ Keywords: Poly-LWE, Ring-LWE, Wide Error distribution, Subset quadruple attack, Sublattice pair with an ideal.

In this paper, we consider the structure of isogeny graphs in SIDH, that is an isogeny-based key-exchange protocol. SIDH is the underlying protocol of SIKE, which is one of the candidates for NIST post quantum cryptography standardization. Since the security of SIDH is based on the hardness of the path-finding problem in isogeny graphs, it is important to study those structure. The existence of cycles in isogeny graph is related to the path-finding problem, so we investigate cycles in the graphs used in SIKE. In particular, we focus on SIKEp434 and SIKEp503, which are the parameter sets of SIKE claimed to satisfy the NIST security level 1 and 2, respectively. We show that there are two cycles in the 3-isogeny graph in SIKEp434, and there is no cycles in the other graphs in SIKEp434 and SIKEp503.

We continue the study of Mastrovito form of Karatsuba multipliers under the shifted polynomial basis (SPB), recently introduced by Li et al. (IEEE TC (2017)). A Mastrovito-Karatsuba (MK) multiplier utilizes the Karatsuba algorithm (KA) to optimize polynomial multiplication and the Mastrovito approach to combine it with the modular reduction. The authors developed a MK multiplier for all trinomials, which obtain a better space and time trade-off compared with previous non-recursive Karatsuba counterparts. Based on this work, we make two types of contributions in our paper. FORMULATION. We derive a new modular reduction formulation for constructing Mastrovito matrix associated with Type II pentanomial. This formula can also be applied to other special type of pentanomials, e.g. Type I pentanomial and Type C.1 pentanomial. Through related formulations, we demonstrate that Type I pentanomial is less efficient than Type II one because of a more complicated modular reduction under the same SPB; conversely, Type C.1 pentanomial is as good as Type II pentanomial under an alternative generalized polynomial basis (GPB). EXTENSION. We introduce a new MK multiplier for Type II pentanomial. It is shown that our proposal is only one $T_X$ slower than the fastest bit-parallel multipliers for Type II pentanomial, but its space complexity is roughly 3/4 of those schemes, where $T_X$ is the delay of one 2-input XOR gate. To the best of our knowledge, it is the first time for hybrid multiplier to achieve such a time delay bound.

The Montgomery ladder and Joye ladder are well-known algorithms for elliptic curve scalar multiplication with a regular structure. The Montgomery ladder is best known for its implementation on Montgomery curves, which requires $5\mathbf{M} + 4\mathbf{S} + 1\mathbf{m} + 8\mathbf{A}$ per scalar bit, and 6 field registers. Here $(\mathbf{M},\mathbf{S},\mathbf{m},\mathbf{A})$ represent respectively field $\mathbf{M}$ultiplications, $\mathbf{S}$quarings, $\mathbf{m}$ultiplications by a curve constant, and $\mathbf{A}$dditions or subtractions. This ladder is also complete, meaning that it works on all input points and all scalars. Many protocols do not use Montgomery curves, but instead use prime-order curves in short Weierstrass form. These have historically been much slower, with ladders costing at least 14 multiplications or squarings per bit: $8\mathbf{M}+6\mathbf{S}+27\mathbf{A}$ for the Montgomery ladder and $8\mathbf{M}+6\mathbf{S}+30\mathbf{A}$ for the Joye ladder. In 2017, Kim et al. improved the Montgomery ladder to $8\mathbf{M}+4\mathbf{S}+12\mathbf{A}+1\mathbf{H}$ per bit using 9 registers, where the $\mathbf{H}$ represents a halving. Hamburg simplified Kim et al.'s formulas to $8\mathbf{M}+4\mathbf{S}+8\mathbf{A}+1\mathbf{H}$ per bit using 6 registers. Here we present improved formulas which compute the Montgomery ladder on short Weierstrass curves using $8\mathbf{M}+3\mathbf{S}+7\mathbf{A}$ per bit, and requiring 6 registers. We also give formulas for the Joye ladder that use $9\mathbf{M}+3\mathbf{S}+7\mathbf{A}$ per bit, requiring 5 registers. One of our new formulas supports very efficient 4-way vectorization. We also discuss curve invariants, exceptional points, side-channel protection and how to set up and finish these ladder operations. Finally, we show a novel technique to make these ladders complete when the curve order is not divisible by 2 or 3, at a modest increase in cost. A sample implementation of these techniques is given in the supplementary material, also posted at https://github.com/bitwiseshiftleft/ladder_formulas.

Deep Learning based Side-Channel Attacks (DL-SCA) are an emerging security assessment method increasingly being adopted by the majority of certification schemes and certification bodies to assess the resistance of cryptographic implementations. The related published investigations have demonstrated that DL-SCA are very efficient when targeting cryptographic designs protected with the common side-channel countermeasures. Furthermore, these attacks allow to streamline the evaluation process as the pre-processing of the traces (\emph{e.g.} alignment, dimensionality reduction, \dots) is no longer mandatory. In practice, the DL-SCA are applied following the divide-and-conquer strategy such that the target, for the training and the attack phases, only depends on $8$ key bits at most (to avoid high time complexity especially during the training). Then, the same process is repeated to recover the remaining bits of the key. To mitigate this practical issue, we propose in this work a new profiling methodology for DL-SCA based on the so-called multi-label classification. We argue that our new profiling methodology allows applying DL-SCA to target a bigger chunk of the key (typically $16$ bits) without introducing a learning time overhead and while guaranteeing a similar attack efficiency compared to the commonly used training strategy. As a side benefit, we demonstrate that our leaning strategy can be applied as well to train several intermediate operations at once. Interestingly, we show that, in this context, our methodology is even faster than the classical training and leads to a more efficient key recovery phase. We validated the soundness of our proposal on simulated traces and experimental data-sets; amongst them, some are publicly available side-channel databases. The obtained results have proven that our profiling methodology is of great practical interest especially in the context of performing penetration tests with high attack potential (\emph{e.g.} Common Criteria, EMVCO) where the time required to perform the attack has an impact on its final rating.

This paper presents WAGE, a new lightweight sponge-based authenticated cipher whose underlying permutation is based on a 37-stage Galois NLFSR over $\mathbb{F}_{2^7}$. At its core, the round function of the permutation consists of the well-analyzed Welch-Gong permutation (WGP), primitive feedback polynomial, a newly designed 7-bit SB sbox and partial word-wise XORs. The construction of the permutation is carried out such that the design of individual components is highly coupled with cryptanalysis and hardware efficiency. As such, we analyze the security of WAGE against differential, linear, algebraic and meet/miss-in-the-middle attacks. For 128-bit authenticated encryption security, WAGE achieves a throughput of 535 Mbps with hardware area of 2540 GE in ASIC ST Micro 90 nm standard cell library. Additionally, WAGE is designed with a twist where its underlying permutation can be efficiently turned into a pseudorandom bit generator based on the WG transformation (WG-PRBG) whose output bits have theoretically proved randomness properties.

In this paper, we present an instruction set coprocessor architecture for lattice-based cryptography and implement the module lattice-based post-quantum key encapsulation mechanism (KEM) Saber as a case study. To achieve fast computation time, the architecture is fully implemented in hardware, including CCA transformations. Since polynomial multiplication plays a performance-critical role in the module and ideal lattice-based public-key cryptography, a parallel polynomial multiplier architecture is proposed that overcomes memory access bottlenecks and results in a highly parallel yet simple and easy-to-scale design. Such multipliers can compute a full multiplication in $256$ cycles, but are designed to target any area/performance trade-offs. Besides optimizing polynomial multiplication, we make important design decisions and perform architectural optimizations to reduce the overall cycle counts as well as improve resource utilization. For the module dimension 3 (security comparable to AES-192), the coprocessor computes CCA key generation, encapsulation, and decapsulation in only 5,453, 6,618 and 8,034 cycles respectively, making it the fastest hardware implementation of Saber to our knowledge. On a Xilinx UltraScale+ XCZU9EG-2FFVB1156 FPGA, the entire instruction set coprocessor architecture runs at 250 MHz clock frequency and consumes 23,686 LUTs, 9,805 FFs, and 2 BRAM tiles (including 5,113 LUTs and 3,068 FFs for the Keccak core).

We facilitate trusted cross-blockchain state proofs by implementing a chain-relay that validates block headers from proof-of-work blockchains. While current approaches require proof sizes linear to the amount of blocks the state was built on, trusted intermediaries, or economic assumptions, we propose the utilization of off-chain computations through zkSNARKs to provide a cryptographically secure and highly scalable sidechain mechanism. Multiple block headers are included in batches and verified off-chain, while preserving light client support. Only the validity of the off-chain computation is verified on-chain, creating a sidechain mechanism that requires constant verification costs and releases the target ledger from processing and storing every single block header of the source blockchain. We provide a prototypical implementation that facilitates the verification of 504 Bitcoin headers in a single proof on Ethereum using the ZoKrates framework. Hereby, the verification costs are reduced by a factor of 187 compared to current approaches such as BTC Relay.

At EUROCRYPT 2004, Naccache et al. showed that the projective coordinates representation of the resulting point of an elliptic curve scalar multiplication potentially allows to recover some bits of the scalar. However, this attack has received little attention by the scientific community, and the status of deployed mitigations to prevent it in widely adopted cryptography libraries is unknown. In this paper, we aim to fill this gap, by analyzing several cryptography libraries in this context. To demonstrate the applicability of the attack, we use a side-channel attack to exploit this vulnerability within libgcrypt in the context of ECDSA. To the best of our knowledge, this is the first practical attack instance. It targets the insecure binary extended Euclidean algorithm implementation using a microarchitectural side-channel attack that allows recovering the projective representation of the output point of scalar multiplication during ECDSA signature generation. We captured 100k traces to estimate the number of traces an attacker would need to compromise the libgcrypt ECDSA implementation, resulting in less than 2k for commonly used elliptic curve secp256r1, demonstrating the attack feasibility. During exploitation, we found two additional vulnerabilities. However, we remark the purpose of this paper is not merely exploiting a library but about providing an analysis on the projective coordinates vulnerability status in widely deployed open-source libraries, filling a gap between its original description in the academic literature and the adoption of countermeasures to thwart it in real-world applications.

The optimization of the main key compression bottlenecks of the supersingular isogeny key encapsulation mechanism (SIKE) has been a target of research in the last few years. Significant improvements were introduced in the recent works of Costello et al. and Zanon et al. The combination of the techniques in previous works reduced the running time of binary torsion basis generation in decompression by a factor of 29 compared to previous work. On the other hand, generating such a basis still takes almost a million cycles on an Intel Core i5-6267U Skylake. In this paper, we continue the work of Zanon et al. and introduce a technique that drops the complexity of binary torsion basis generation by a factor log p in the number of underlying field multiplications. In particular, our experimental results show that a basis can be generated in about 1,300 cycles, attaining an improvement by a factor more than 600. Although this result eliminates one of the key compression bottlenecks, many other bottlenecks remain. In addition, we give further improvements for the ternary torsion generation with significant impact on the related decompression procedure. Moreover, a new trade-off between ciphertext sizes vs decapsulation speed and storage is introduced and achieves a 1.7 times faster decapsulation.

The one-out-of-many proof is a cryptographic zero-knowledge construction enabling the prover to demonstrate knowledge of a secret element among the given public list of cryptographic commitments opening to zero. This method is relying on standard Decisional Diffie-Hellman security assumptions and can result in efficient accountable ring signature schemes [4] and proofs of set memberships [5] with a signature size smaller than all existing alternative schemes relying on standard assumptions. This construction also serves as a fundamental building block for numerous recent blockchain privacy protocols including Anonymous Zether, Zerocoin, Lelantus, Lelantus-MW, Triptych and Triptych-2. One-out-of-many proofs require O(logN)-sized communication and can be implemented in O(N) time for the verifier and O(NlogN) time for the prover. In this work, we introduce a new method of instantiating one-out-of-many proofs which reduces the proof generation time by an order of magnitude. In certain practical applications our method also helps to fasten the verification process of multiple simultaneously generated proofs. Our approach still results in shorter proofs comprised of only a logarithmic number of commitments and does not compromise the highly efficient batch verification properties endemic to the original construction. We believe this work can also foster further research towards building more efficient one-out-of-many proofs which are extremely useful constructions in the blockchain privacy space and beyond.

Mathematics and cryptography have a long history together, with the ups and downs inherent in any long relationship. Whether it is a marriage of convenience or a love match, their progeny have lives of their own and have had an impact on the world. This invited lecture will briefly recall some high points from the past, give speculation and encouragement for the future of this marriage, and give counseling on how to improve communication, resolve conflicts, and play well together, based on personal experience and lessons learned.

In a joint effort to fight the COVID-19 pandemic, Apple Inc. and Google Inc. recently partnered to develop a contact tracing technology, inspired by the DP-3T and TCN protocols, to help governments and health agencies reduce the spread of the virus, with user privacy and security central to the design. The partnership announcement included technical specifications of the planned technology, which has great potential for widespread adoption due to the global reach of the two companies. At the same time, the anonymous distributed setting for contact tracing as well as other aspects of the specifications create opportunities for attackers to mount common attacks on the technology. In this work, we provide a security analysis of these specifications, the initial version of which was communicated early on to Apple Inc. in April this year, soon after announcement of the specifications. We show that the current specifications may introduce significant risks to society due to the common attacks and propose novel mitigation strategies for these risks that do not require major changes to the technology and are easy to adopt. To the best of our knowledge, ours is the first contact tracing proposal to mitigate the risks of all these common attacks in the anonymous distributed setting without introducing architectural changes. Our analysis focuses mostly on system security considerations, which have not been well covered previously, yet also includes novel information security considerations. We leave out of scope a discussion on how important or effective the technology is in fighting the pandemic.

Picnic is a digital signature algorithm designed to provide security against attacks by quantum computers. The design uses only symmetric-key primitives, and is an efficient instantiation of the MPC-in-the-head paradigm. In this work, we explore the Picnic design in great detail. We investigate and benchmark different parameter choices and show that there exist better parameter choices than those in the current specification. We also present improvements to the MPC protocol that shorten signatures and reduce signing time. The proposed MPC changes tailor the protocol to the circuit of interest in Picnic, but may also be of independent interest. Taken together, these changes give a new instantiation of Picnic that signs messages 7.9 to 13.9 times faster, and verifies signatures 4.5 to 5.5 times faster than the existing ``Picnic2'' design, while having nearly the same signature sizes.

The COVID-19 pandemic has posed a unique challenge for the world to find solutions, ranging from vaccines to ICT solutions to slow down the virus spreading. Due to the highly contagious nature of the virus, social distancing is one fundamental measure which has already adopted by many countries. At the technical level, this prioritises contact tracing solutions, which can alert the users who have been in close contact with the infected persons and meanwhile allow heath authorities to take proper actions. In this paper, we examine several existing privacy-aware contact tracing solutions and analyse their (dis)advantages. At the end, we describe several major observations and outline an interdisciplinary research agenda towards more comprehensive and effective privacy-aware contact tracing solutions.

In this work we present a duplex-based authenticated encryption scheme Friet based on a new permutation called Friet-P. We designed Friet-P with a novel approach for cryptographic permutations and block ciphers that takes fault-attack resistance into account and that we introduce in this paper. In this method, we build a permutation $f_C$ to be embedded in a larger one, $f$ . First, we define $f$ as a sequence of steps that all abide a chosen error-correcting code $C$, i.e., that map $C$-codewords to $C$-codewords. Then, we embed $f_C$ in $f$ by first encoding its input to an element of $C$, applying $f$ and then decoding back from $C$. This last step detects a fault when the output of $f$ is not in $C$. We motivate the design of the permutation we use in Friet and report on performance in soft- and hardware. We evaluate the fault-detection capabilities of the software and simulated hardware implementations with attacks. Finally, we perform a leakage evaluation. Our code is available at https://github.com/thisimon/Friet.git.

The golden collision problem asks us to find a single, special collision among the outputs of a pseudorandom function. This generalizes meet-in-the-middle problems, and is thus applicable in many contexts, such as cryptanalysis of the NIST post-quantum candidate SIKE. The main quantum algorithms for this problem are memory-intensive, and the costs of quantum memory may be very high. The quantum circuit model implies a linear cost for random access, which annihilates the exponential advantage of the previous quantum collision-finding algorithms over Grover's algorithm or classical van Oorschot-Wiener. Assuming that quantum memory is costly to access but free to maintain, we provide new quantum algorithms for the golden collision problem with high memory requirements but low gate costs. Under the assumption of a two-dimensional connectivity layout, we provide better quantum parallelization methods for generic and golden collision finding. This lowers the quantum security of the golden collision and meet-in-the-middle problems, including SIKE.

We prove that the equivalence of two fundamental problems in the theory of computing. For every polynomial $t(n)\geq (1+\varepsilon)n, \varepsilon>0$, the following are equivalent: - One-way functions exists (which in turn is equivalent to the existence of secure private-key encryption schemes, digital signatures, pseudorandom generators, pseudorandom functions, commitment schemes, and more); - $t$-time bounded Kolmogorov Complexity, $K^t$, is mildly hard-on-average (i.e., there exists a polynomial $p(n)>0$ such that no $\PPT$ algorithm can compute $K^t$, for more than a $1-\frac{1}{p(n)}$ fraction of $n$-bit strings). In doing so, we present the first natural, and well-studied, computational problem characterizing the feasibility of the central private-key primitives and protocols in Cryptography.

Pattern matching is one of the most fundamental and important paradigms in several application domains such as digital forensics, cyber threat intelligence, or genomic and medical data analysis. While it is a straightforward operation when performed on plaintext data, it becomes a challenging task when the privacy of both the analyzed data and the analysis patterns must be preserved. In this paper, we propose new provably correct, secure, and relatively efficient (compared to similar existing schemes) public and private key based constructions that allow arbitrary pattern matching over encrypted data while protecting both the data to be analyzed and the patterns to be matched. That is, except the pattern provider (resp. the data owner), all other involved parties in the proposed constructions will learn nothing about the patterns to be searched (resp. the data to be inspected). Compared to existing solutions, the constructions we propose has some interesting properties: (1) the size of the ciphertext is linear to the size of plaintext and independent of the sizes and the number of the analysis patterns; (2) the sizes of the issued trapdoors are constant on the size of the data to be analyzed; and (3) the search complexity is linear on the size of the data to be inspected and is constant on the sizes of the analysis patterns. The conducted evaluations show that our constructions drastically improve the performance of the most efficient state of the art solution.

Blockchain Sharding is a blockchain performance enhancement approach. By splitting a blockchain into several parallel-run committees (shards), it helps increase transaction throughput, reduce resources required, and increase reward expectation for participants. Recently, several flexible sharding methods that can tolerate up to $n/2$ Byzantine nodes ($n/2$ security level) have been proposed. However, these methods suffer from two main drawbacks. First, in a non-sharding blockchain, nodes can have different weight (power or stake) to create a consensus. So an adversary needs to control half of the overall weight of the system in order for a piece of faulty information to be accepted into the blockchain ($p/2$ security level). In blockchain sharding, all nodes carry the same weight. Thus, it is only under the assumption that the honest participants are creating as many nodes as they can that a $n/2$ security level blockchain sharding reaches the $p/2$ security level. Secondly, when some nodes leave the system, other nodes need to be reassigned, frequently, from shard to shard in order to maintain the security level of the system. In this paper, we present Multichain MWPoW, a $p/2$ security level blockchain sharding architecture that does not require honest participants to create multiple nodes and requires less node reassignment when some nodes leave the system. It combines the Multiple Winners Proof of Work consensus protocol (MWPoW) with the flexibility of $n/2$ blockchain sharding. Our experiments show that Multichain MWPoW outperforms existing blockchain sharding approaches in terms of security, transaction throughput and flexibility.

We present a card-based protocol for computing a three-input majority using six cards. The protocol essentially consists of performing a simple XOR protocol two times. Compared to the existing protocols, our protocol does not require private operations other than choosing cards.

Vector commitments enable a user to commit to a sequence of values and provably reveal one or many values at specific positions at a later time. In this work, we construct Pointproofs--a new vector commitment scheme that supports non-interactive aggregation of proofs across multiple commitments. Our construction enables any third party to aggregate a collection of proofs with respect to different, independently computed commitments into a single proof represented by an elliptic curve point of 48-bytes. In addition, our scheme is hiding: a commitment and proofs for some values reveal no information about the remaining values. We build Pointproofs and demonstrate how to apply them to blockchain smart contracts. In our example application, Pointproofs reduce bandwidth overheads for propagating a block of transactions by at least 60% compared to prior state-of-art vector commitments. Pointproofs are also efficient: on a single-thread, it takes 0.08 seconds to generate a proof for 8 values with respect to one commitment, 0.25 seconds to aggregate 4000 such proofs across multiple commitments into one proof, and 23 seconds (0.7 ms per value proven) to verify the aggregated proof.

Currently several projects (including DP-3T, east and west coast PACT, Covid watch) aim at designing and implementing protocols for privacy preserving automated contact tracing to help fight the current pandemic. Those proposal are very similar, and in their most basic from basically propose an app for mobile phones which broadcasts frequently changing pseudorandom identifiers via (low energy) Bluetooth, and at the same time, the app stores IDs broadcast by phones in its proximity. Only if a user is tested positive, their IDs of the last 14 days are published so other users can check if they have stored them locally and thus were close to an infected person. Vaudenay [eprint 2020/399] observes that this basic scheme (he considers the DP-3T proposal) succumbs to relay and even replay attacks, and proposes more complex \emph{interactive} schemes which prevent those attacks without giving up too many privacy aspects. Unfortunately interaction is problematic for this application for efficiency and security reasons. In this note propose a simple \emph{non-interactive} variant of the basic protocol that \begin{itemize} \item (security) Provably prevents replay and relay attacks. \item (privacy) The data of all parties (even jointly) reveals no information on the location or time where encounters happened. \item (efficiency) The broadcasted message can fit into 128 bits and uses simple crypto (commitments and secret key authentication). \end{itemize} Towards this end we introduce the concept of ``delayed authentication", which basically is a message authentication code where verification can be done in two steps, where the first doesn't require the key, and the second doesn't require the message.

Since its proposal in Asiacrypt 2018, the commutative isogeny-based key exchange protocol (CSIDH) has spurred considerable attention to improving its performance and re-evaluating its classical and quantum security guarantees. In this paper we discuss how the optimal strategies employed by the Supersingular Isogeny Diffie-Hellman (SIDH) key agreement protocol can be naturally extended to CSIDH. Furthermore, we report a software library that achieves moderate but noticeable performance speedups when compared against state-of-the-art implementations of CSIDH-512, which is the most popular CSIDH instantiation. We also report an estimated number of field operations for larger instantiations of this protocol, namely, CSIDH-1024 and CSIDH-1792.

We introduce the Multi-Base Discrete Logarithm (MBDL) problem. We use this to give reductions, for Schnorr and Okamoto identification and signatures, that are non-rewinding and, by avoiding the notorious square-root loss, tighter than the classical ones from the Discrete Logarithm (DL) problem. This fills a well-known theoretical and practical gap regarding the security of these schemes. We show that not only is the MBDL problem hard in the generic group model, but with a bound that matches that for DL, so that our new reductions justify the security of these primitives for group sizes in actual use.

Candidates of Indistinguishability Obfuscation (iO) can be categorized as ``direct'' or ``bootstrapping based''. Direct constructions rely on high degree multilinear maps [GGH13,GGHRSW13] and provide heuristic guarantees, while bootstrapping based constructions [LV16,Lin17,LT17,AJLMS19,Agr19,JLMS19] rely, in the best case, on bilinear maps as well as new variants of the Learning With Errors (LWE) assumption and pseudorandom generators. Recent times have seen exciting progress in the construction of indistinguishability obfuscation (iO) from bilinear maps (along with other assumptions) [LT17,AJLMS19,JLMS19,Agr19]. As a notable exception, a recent work by Agrawal [Agr19] provided a construction for iO without using any maps. This work identified a new primitive, called Noisy Linear Functional Encryption (NLinFE) that provably suffices for iO and gave a direct construction of NLinFE from new assumptions on lattices. While a preliminary cryptanalysis for the new assumptions was provided in the original work, the author admitted the necessity of performing significantly more cryptanalysis before faith could be placed in the security of the scheme. Moreover, the author did not suggest concrete parameters for the construction. In this work, we fill this gap by undertaking the task of thorough cryptanalytic study of NLinFE. We design two attacks that let the adversary completely break the security of the scheme. To achieve this, we develop new cryptanalytic techniques which (we hope) will inform future designs of the primitive of NLinFE. From the knowledge gained by our cryptanalytic study, we suggest modifications to the scheme. We provide a new scheme which overcomes the vulnerabilities identified before. We also provide a thorough analysis of all the security aspects of this scheme and argue why plausible attacks do not work. We additionally provide concrete parameters with which the scheme may be instantiated. We believe the security of NLinFE stands on significantly firmer footing as a result of this work.

Quantum money allows a bank to mint quantum money states that can later be verified and cannot be forged. Usually, this requires a quantum communication infrastructure to transfer quantum states between the user and the bank. Gavinsky (CCC 2012) introduced the notion of classically verifiable quantum money, which allows verification through classical communication. In this work we introduce the notion of classical minting, and combine it with classical verification to introduce semi-quantum money. Semi-quantum money is the first type of quantum money to allow transactions with completely classical communication and an entirely classical bank. This work features constructions for both a public memory-dependent semi-quantum money scheme and a private memoryless semi-quantum money scheme. The public construction is based on the works of Zhandry and Coladangelo, and the private construction is based on the notion of Noisy Trapdoor Claw Free Functions (NTCF) introduced by Brakerski et al. (FOCS 2018). In terms of technique, our main contribution is a perfect parallel repetition theorem for NTCF.

The goal of white-box cryptography is to protect secret keys embedded in a cryptographic software deployed in an untrusted environment. In this article, we revisit state-of-the-art countermeasures employed in white-box cryptography, and we discuss possible ways to combine them. Then we analyze the different gray-box attack paths and study their performances in terms of required traces and computation time. Afterward, we propose a new paradigm for the gray-box attack against white-box cryptography, which exploits the data-dependency of the target implementation. We demonstrate that our approach provides substantial complexity improvements over the existing attacks. Finally, we showcase this new technique by breaking the three winning AES-128 white-box implementations from WhibOx 2019 white-box cryptography competition.

The GIFT family of lightweight block ciphers, published at CHES 2017, offers excellent hardware performance figures and has been used, in full or in part, in several candidates of the ongoing NIST lightweight cryptography competition. However, implementation of GIFT in software seems complex and not efficient due to the bit permutation composing its linear layer (a feature shared with PRESENT cipher). In this article, we exhibit a new non-trivial representation of the GIFT family of block ciphers over several rounds. This new representation, that we call fixslicing, allows extremely efficient software bitsliced implementations of GIFT, using only a few rotations, surprisingly placing GIFT as a very efficient candidate on micro-controllers. Our constant time implementations show that, on ARM Cortex-M3, 128-bit data can be ciphered with only about 800 cycles for GIFT-64 and about 1300 cycles for GIFT-128 (assuming pre-computed round keys). In particular, this is much faster than the impressive PRESENT implementation published at CHES 2017 that requires 2116 cycles in the same setting, or the current best AES constant time implementation reported that requires 1617 cycles. This work impacts GIFT, but also improves software implementations of all other cryptographic primitives directly based on it or strongly related to it.

Secure multi-party computation has been extensively studied in the past years and has reached a level that is considered practical for several applications. The techniques developed thus far have been steadily optimized for performance and were shown to be secure in the classical setting, but are not known to be secure against quantum adversaries. In this work, we start to pave the way for secure two-party computation in a quantum world where the adversary has access to a quantum computer. We show that post-quantum secure two-party computation has comparable efficiency to their classical counterparts. For this, we develop a lattice-based OT protocol which we use to implement a post-quantum secure variant of Yao's famous garbled circuits (GC) protocol (FOCS'82). Along with the OT protocol, we show that the oblivious transfer extension protocol of Ishai et al. (CRYPTO'03), which allows running many OTs using mainly symmetric cryptography, is post-quantum secure. To support these results, we prove that Yao's GC protocol achieves post-quantum security if the underlying building blocks do.

We present the first practical software implementation of Supersingular Isogeny Key Encapsulation (SIKE) round 2, targeting NIST's 1, 2, 3, and 5 security levels on 32-bit ARM Cortex-M4 microcontrollers. The proposed library introduces a new speed record of all SIKE Round 2 protocols with reasonable memory consumption on the low-end target platforms. We achieved this record by adopting several state-of-the-art engineering techniques as well as highly-optimized hand-crafted assembly implementation of finite field arithmetic. In particular, we carefully redesign the previous optimized implementations of finite field arithmetic on the 32-bit ARM Cortex-M4 platform and propose a set of novel techniques which are explicitly suitable for SIKE primes. The benchmark result on STM32F4 Discovery board equipped with 32-bit ARM Cortex-M4 microcontrollers shows that entire key encapsulation and decapsultation over SIKEp434 take about 184 million clock cycles (i.e. 1.09 seconds @168MHz). In contrast to the previous optimized implementation of the isogeny-based key exchange on low-end 32-bit ARM Cortex-M4, our performance evaluation shows feasibility of using SIKE mechanism on the target platform. In comparison to the most of the post-quantum candidates, SIKE requires an excessive number of arithmetic operations, resulting in significantly slower timings. However, its small key size makes this scheme as a promising candidate on low-end microcontrollers in the quantum era by ensuring the lower energy consumption for key transmission than other schemes.

The US National Institute of Standards and Technology (NIST) recently announced the public-key cryptosystems (PKC) that have passed to the second round of the post-quantum standardization process. Most of these PKC come in two flavours: a weak IND-CPA version and a strongly secure IND-CCA construction. For the weaker scheme, no level of security is claimed in the plaintext-checking attack (PCA) model. However, previous works showed that, for several NIST candidates, only a few PCA queries are sufficient to recover the secret key. In order to create a more complete picture, we design new key-recovery PCA against several round 2 candidates. Our attacks against CRYSTALS-Kyber, HQC, LAC and SABER are all practical and require only a few thousand queries to recover the full secret key. In addition, we present another KR-PCA attack against the rank-based scheme RQC, which needs roughly $O(2^{38})$ queries. Hence, this type of scheme seems to resist better than others to key recovery. Motivated by this observation, we prove an interesting result on the rank metric. Namely, that the learning problem with the rank distance is hard for some parameters, thus invalidating a common strategy for reaction attacks.

Rainbow is a signature scheme that is based on multivariate polynomials. It is one of the Round-2 candidates of the NIST’s Post-Quantum Cryptography Standardization project. Its computations rely heavily on GF(2^8) arithmetic and the Rainbow submission optimizes the code by using AVX2 shuffle and permute instructions. In this paper, we show a new optimization that leverages: a) AVX512 architecture; b) the latest processor capabilities Galois Field New Instructions(GF-NI), available on Intel "Ice Lake" processor. We achieved a speedup of 2.43x/3.13x/0.64x for key generation/signing/verifying, respectively. We also propose a variation of Rainbow, with equivalent security, using a different representation of GF(2^8). With this variant, we achieve a speedup of 2.44x/4.7x/2.1x for key generation/signing/verifying, respectively.

Abstract. With the growth of cloud computing, the need arises for Private Set Intersection (PSI) protocols that can operate on outsourced data and delegate computation to cloud servers. One limitation of existing delegated PSI protocols is that they are all designed for static data and do not allow efficient update on outsourced data. Another limitation is that they cannot efficiently support PSI among multiple clients, which is often needed in practice. This paper presents “Feather”, the first delegated PSI protocol that supports efficient data updates and scalable multi-party PSI computation on outsourced datasets. It lets clients independently prepare and upload their private data to the cloud once, then delegate the computation an unlimited number of times. The update operation has very low communication and computation complexity, and this is achieved without sacrificing PSI efficiency and security. Feather does not use public key cryptography, that makes it more scalable. We have implemented a prototype and compared the concrete performance against the state of the art. The evaluation indicates that Feather does achieve better performance in both update and PSI computation.

Optimistic responsiveness was introduced to shorten the latency of a synchronous Byzantine consensus protocol that is inherently lower bounded by the pessimistic bound on the network delay $\Delta$. It states that a protocol makes a decision with latency on the order of actual network delay $ \delta $ if the number of actual faults is significantly smaller than $f$, which is the worst-case allowed. In this paper, we investigate if a Byzantine consensus can simultaneously achieve (i) optimistic responsiveness, and (ii) optimal latency of $\Delta + O(\delta)$ in the presence of $f$ faults. To do this, we provide a tight upper bound on the number of actual faults by showing matching feasibility and infeasibility results. Furthermore, we present a simple leader-based Byzantine fault-tolerant (BFT) replication protocol as a practical application. Even while being able to rotate leaders after every decision, our protocol simultaneously achieves average latency of (i) $3\delta$ under optimistic condition and (ii) $1.5\Delta + O(\delta)$ (or $3\Delta + O(\delta)$) in the presence of $f$ faults, which is more than a factor of two better than current state-of-the-art rotating-leader BFT protocols.

Modern electronic voting systems (e-voting systems) are designed to provide not only vote privacy but also (end-to-end) verifiability. Several verifiable e-voting systems have been proposed in the literature, with Helios being one of the most prominent ones. Almost all such systems, however, reveal not just the voting result but also the full tally, consisting of the exact number of votes per candidate or even all single votes. There are several situations where this is undesirable. For example, in elections with only a few voters (e.g., boardroom or jury votings), revealing the complete tally leads to a low privacy level, possibly deterring voters from voting for their actual preference. In other cases, revealing the complete tally might unnecessarily embarrass some candidates. Often, the voting result merely consists of a single winner or a ranking of candidates, so revealing only this information but not the complete tally is sufficient. This property is called tally-hiding and it offers completely new options for e-voting. In this paper, we propose the first provably secure end-to-end verifiable tally-hiding e-voting system, called Ordinos. We instantiated our system with suitable cryptographic primitives, including an MPC protocol for greater-than tests, implemented the system, and evaluated its performance, demonstrating its practicality. Moreover, our work provides a deeper understanding of tally-hiding in general, in particular in how far tally-hiding affects the levels of privacy and verifiability of e-voting systems.

In this work we develop a rewarding framework that can be used as a building block in crowd-sensing applications. Although a core requirement of such systems is user engagement, people may be reluctant to participate as sensitive information about them may be leaked or inferred from submitted data. Thus monetary incentives could help attract a large number of participants, thereby increasing not only the amount but also the quality of sensed data. Our first contribution in this work is to ensure that users can submit data and obtain Bitcoin payments in a privacy-preserving manner, preventing curious providers from linking the data or the payments back to the user. At the same time, we thwart malicious user behavior such as double-redeeming attempts where a user tries to obtain rewards for multiple submissions of the same data. More importantly, we ensure the fairness of the exchange in a completely trustless manner; by relying on the Blockchain, we eliminate the trust placed on third parties in traditional fair exchange protocols. Finally, our system is highly efficient as most of the protocol steps do not utilize the Blockchain network. When they do, we only rely on simple Bitcoin transactions as opposed to prior works that are based on the use of highly complex smart contracts.

Chameleon-hash functions, introduced by Krawczyk and Rabin at NDSS 2000, are trapdoor collision-resistant hash-functions parametrized by a public key. If the corresponding secret key is known, arbitrary collisions for the hash function can be efficiently found. Chameleon-hash functions have prominent applications in the design of cryptographic primitives, such as lifting non-adaptively secure signatures to adaptively secure ones. Recently, this primitive also received a lot of attention as a building block in more complex cryptographic applications ranging from editable blockchains to advanced signature and encryption schemes. We observe that in latter applications various different notions of collision-resistance are used, and it is not always clear if the respective notion does really cover what seems intuitively required by the application. Therefore, we revisit existing collision-resistance notions in the literature, study their relations, and - using the example of the recent redactable blockchain proposals - discuss which practical impact different notions of collision-resistance might have. Moreover, we provide a stronger, and arguably more desirable, notion of collision-resistance than what is known from the literature. Finally, we present a surprisingly simple and efficient black-box construction of chameleon-hash functions achieving this strong notion.

In this short note, we show that substantially weaker Low Order assumptions are sufficient to prove the soundness of Pietrzak's protocol for proof of exponentiation in groups of unknown order. This constitutes the first step to a better understanding of the asymptotic computational complexity of breaking the soundness of the protocol. Furthermore, we prove the equivalence of the (weaker) Low Order assumption(s) and the Factoring assumption in RSA groups for a non-negligible portion of moduli. We argue that in practice our reduction applies for a considerable amount of deployed moduli. Our results have cryptographic applications, most importantly in the theory of recently proposed verifiable delay function constructions. Finally, we describe how to certify RSA moduli free of low order elements.

Non-interactive zero-knowledge proofs, and more specifically succinct non-interactive zero-knowledge arguments (zk-SNARKs), have been proven to be the “swiss army knife” of the blockchain and distributed ledger space, with a variety of applications in privacy, interoperability and scalability. Many commonly used SNARK systems rely on a structured reference string, the secure generation of which turns out to be their Achilles heel: If the randomness used for the generation is known, the soundness of the proof system can be broken with devastating consequences for the underlying blockchain system that utilises them. In this work we describe and analyze, for the first time, a blockchain mechanism that produces a secure SRS with the characteristic that security is shown for the exact same conditions under which the blockchain protocol is proven to be secure. Our mechanism makes use of the recent discovery of updateable structure reference strings to perform this secure generation in a fully distributed manner. In this way, the SRS emanates from the normal operation of the blockchain protocol itself without the need of additional security assumptions or off-chain computation and/or verification. We provide concrete guidelines for the parameterisation of this system which allows for the completion of a secure setup in a reasonable period of time. We also provide an incentive scheme that, when paired with the update mechanism, properly incentivises participants into contributing to secure reference string generation.

In an article presented at FDTC 2018, Arribas, De Cnudde, and Sijacic prove under mild conditions that threshold implementations (TIs) are secure against fault sensitivity analysis (FSA). Later in 2018, in the PhD thesis of De Cnudde, additional assumptions were imposed to provably withstand FSA, thereby increasing the required number of random bits. We point out that even under the latter, stronger conditions, the proof is incorrect.

To help fighting the COVID-19 pandemic, the Pan-European Privacy-Preserving Proximity Tracing (PEPP-PT) project proposed a Decentralized Privacy-Preserving Proximity Tracing (DP3T) system. This helps tracking the spread of SARS-CoV-2 virus while keeping the privacy of individuals safe. In this report, we analyze the security and the privacy protection of DP3T. Without questioning how effective it could be against the pandemic, we show that it may introduce severe risks to society. Furthermore, we argue that some privacy protection measurements by DP3T may have the opposite affect of what they were intended to. Specifically, sick and reported people may be deanonymized, private encounters may be revealed, and people may be coerced to reveal the private data they collect.

Contact tracing is a promising approach to combat the COVID-19 pandemic. Various systems have been proposed to automatise the process. Many designs rely heavily on a centralised server or reveal significant amounts of private data to health authorities. We propose CAUDHT, a decentralized peer-to-peer system for contact tracing. The central health authority can focus on providing and operating tests for the disease while contact tracing is done by the system’s users themselves. We use a distributed hash table to build a decentral messaging system for infected patients and their contacts. With blind signatures, we ensure that messages about infections are authentic and unchanged. A strong privacy focus enables data integrity, confidentiality, and privacy.

In this paper, we present all 4-bit S-boxes which are able to support BOGI logic. We exhaustively show that only 2,413 PXE classes of 4-bit S-box are BOGI-applicable among the 142,090,700 PXE classes. We evaluate the whole BOGI-applicable S-boxes in terms of the security and implementation costs. The security evaluation includes security strength of the S-boxes themselves, and how they affect the resistance of GIFT-64 against differential and linear cryptanalysis (DC and LC). The security evaluation shows that all the BOGI-applicable S-boxes fulfill the security criteria of GIFT designers as long as they have the differential uniformity and linearity as 6 and 8, respectively. It will also be shown that the security of GIFT-64 against DC and LC can be improved only by changing the S-box. Moreover, we evaluate the implementation costs of the BOGI-applicable S-boxes by finding their optimal implementation. The results show that GIFT S-box is well-chosen considering existence of fixed-points, and suggest a set of S-boxes providing the same implementation cost as GIFT S-box. Finally, we suggest a set of potentially better S-boxes for GIFT-64 based on our investigations.

In recent years, deep learning-based side-channel attacks have established their position as mainstream. However, most deep learning techniques for cryptanalysis mainly focused on classifying side-channel information in a profiled scenario where attackers can obtain a label of training data. In this paper, we introduce a novel approach with deep learning for improving side-channel attacks, especially in a non-profiling scenario. We also propose a new principle of training that trains an autoencoder through the noise from real data using noise-reduced labels. It notably diminishes the noise in measurements by modifying the autoencoder framework to the signal preprocessing. We present convincing comparisons on our custom dataset, captured from ChipWhisperer-Lite board, that demonstrate our approach outperforms conventional preprocessing methods such as principal component analysis and linear discriminant analysis. Furthermore, we apply the proposed methodology to realign de-synchronized traces that applied hiding countermeasures, and we experimentally validate the performance of the proposal. Finally, we experimentally show that we can improve the performance of higher-order side-channel attacks by using the proposed technique with domain knowledge for masking countermeasures.

Reductions between problems, the mainstay of theoretical computer science, efficiently map an instance of one problem to an instance of another in such a way that solving the latter allows solving the former. The subject of this work is ``lossy'' reductions, where the reduction loses some information about the input instance. We show that such reductions, when they exist, have interesting and powerful consequences for lifting hardness into ``useful'' hardness, namely cryptography. Our first, conceptual, contribution is a definition of lossy reductions in the language of mutual information. Roughly speaking, our definition says that a reduction $\mathsf{C}$ is $t$-lossy if, for any distribution $X$ over its inputs, the mutual information $I(X;\mathsf{C}(X)) \leq t$. Our treatment generalizes a variety of seemingly related but distinct notions such as worst-case to average-case reductions, randomized encodings (Ishai and Kushilevitz, FOCS 2000), homomorphic computations (Gentry, STOC 2009), and instance compression (Harnik and Naor, FOCS 2006). We then proceed to show several consequences of lossy reductions: 1. We say that a language $L$ has an $f$-reduction to a language $L'$ for a Boolean function $f$ if there is a (randomized) polynomial-time algorithm $\mathsf{C}$ that takes an $m$-tuple of strings $X = (x_1,\ldots,x_m)$, with each $x_i\in\{0,1\}^n$, and outputs a string $z$ such that with high probability, \begin{align*} L'(z) = f(L(x_1),L(x_2),\ldots,L(x_m)) \end{align*} 2. Suppose a language $L$ has an $f$-reduction $\mathsf{C}$ to $L'$ that is $t$-lossy. Our first result is that one-way functions exist if $L$ is worst-case hard and one of the following conditions holds: - $f$ is the OR function, $t \leq m/100$, and $L'$ is the same as $L$ - $f$ is the Majority function, and $t \leq m/100$ - $f$ is the OR function, $t \leq O(m\log{n})$, and the reduction has no error This improves on the implications that follow from combining (Drucker, FOCS 2012) with (Ostrovsky and Wigderson, ISTCS 1993) that result in {\em auxiliary-input} one-way functions. 3. Our second result is about the stronger notion of $t$-compressing $f$-reductions -- reductions that only output $t$ bits. We show that if there is an average-case hard language $L$ that has a $t$-compressing Majority reduction to some language for $t=m/100$, then there exist collision-resistant hash functions. This improves on the result of (Harnik and Naor, STOC 2006), whose starting point is a cryptographic primitive (namely, one-way functions) rather than average-case hardness, and whose assumption is a compressing OR-reduction of SAT (which is now known to be false unless the polynomial hierarchy collapses). Along the way, we define a non-standard one-sided notion of average-case hardness, which is the notion of hardness used in the second result above, that may be of independent interest.

We propose a new approach to construct general-purpose indistinguishability obfuscation (iO). Our construction is obtained via a new intermediate primitive that we call split fully-homomorphic encryption (split FHE), which we show to be sufficient for constructing iO. Specifically, split FHE is FHE where decryption takes the following two-step syntactic form: (i) A secret decryption step uses the secret key and produces a hint which is (asymptotically) shorter than the length of the encrypted message, and (ii) a public decryption step that only requires the ciphertext and the previously generated hint (and not the entire secret key), and recovers the encrypted message. In terms of security, the hints for a set of ciphertexts should not allow one to violate semantic security for any other ciphertexts. Next, we show a generic candidate construction of split FHE based on three building blocks: (i) A standard FHE scheme with linear decrypt-and-multiply (which can be instantiated with essentially all LWE-based constructions), (ii) a linearly homomorphic encryption scheme with short decryption hints (such as the Damgard-Jurik encryption scheme, based on the DCR problem), and (iii) a cryptographic hash function (which can be based on a variety of standard assumptions). Our approach is heuristic in the sense that our construction is not provably secure and makes implicit assumptions about the interplay between these underlying primitives. We show evidence that this construction is secure by providing an argument in an appropriately defined oracle model. We view our construction as a big departure from the state-of-the-art constructions, and it is in fact quite simple.

We study the problem of secure two-party computation of arithmetic circuits in the presence of active (``malicious'') parties. This problem is motivated by privacy-preserving numerical computations, such as ones arising in the context of machine learning training and classification, as well as in threshold cryptographic schemes. In this work, we design, optimize, and implement an actively secure protocol for secure two-party arithmetic computation. A distinctive feature of our protocol is that it can make a fully modular black-box use of any passively secure implementation of oblivious linear function evaluation (OLE). OLE is a commonly used primitive for secure arithmetic computation, analogously to the role of oblivious transfer in secure computation for Boolean circuits. For typical (large but not-too-narrow) circuits, our protocol requires roughly 4 invocations of passively secure OLE per multiplication gate. This significantly improves over the recent TinyOLE protocol (Dottling et al., ACM CCS 2017), which requires 22 invocations of actively secure OLE in general, or 44 invocations of a specific code-based passively secure OLE. Our protocol follows the high level approach of the IPS compiler (Ishai et al., CRYPTO 2008, TCC 2009), optimizing it in several ways. In particular, we adapt optimization ideas that were used in the context of the practical zero-knowledge argument system Ligero (Ames et al., ACM CCS 2017) to the more general setting of secure computation, and explore the possibility of boosting efficiency by employing a ``leaky'' passively secure OLE protocol. The latter is motivated by recent (passively secure) lattice-based OLE implementations in which allowing such leakage enables better efficiency. We showcase the performance of our protocol by applying its implementation to several useful instances of secure arithmetic computation. On ``wide'' circuits, such as ones computing a fixed function on many different inputs, our protocol is 5x faster and transmits 4x less data than the state-of-the-art Overdrive (Keller et al., Eurocrypt 2018). Our benchmarks include a general passive-to-active OLE compiler, authenticated generation of ``Beaver triples'', and a system for securely outsourcing neural network classification. The latter is the first actively secure implementation of its kind, strengthening the passive security provided by recent related works (Mohassel and Zhang, IEEE S&P 2017; Juvekar et al., USENIX 2018).

LRBC is a new lightweight block cipher that has been proposed for resource-constrained IoT devices. The cipher is claimed to be secure against differential cryptanalysis and linear cryptanalysis. However, beside short state length which is only 16-bits, the structures of the cipher only use the linear operations, the its s-boxes, and this is a reason why the cipher is completely insecure against the mentioned attacks. we present a few examples to show that. Also, we show that the round function of LRBC has some structural problem and even if we fix them the cipher does not provide complete diffusion. Hence, even with replacement of the cipher s-boxes with proper s-boxes, the problem will not be fixed and it is possible to provide deterministic distinguisher for any number of round of the cipher. In addition, we show that for any fixed key, it is possible to create a full code book for the cipher with the complexity of $2^{n/2}$, which should be compared with $2^{n}$ for any secure $n$-bit block cipher.

The implementation of isogeny-based cryptography mainly use Montgomery curves as they offer fast elliptic curve arithmetic and isogeny compuation. However, although Montgomery curves have efficient 3- and 4-isogenies, it becomes inefficient when recovering the coefficient of the image curve for large degree isogenies. This is the main bottleneck of using a Montgomery curve for CSIDH as it requires odd-degree isogenies up to at least 587. In this paper, we present a new optimization method for faster CSIDH protocols entirely on Montgomery curves. To this end, we present a new parameter for CSIDH in which the rational 2-torsion points are defined over $\mathbb{F}_p$. By using the proposed parameters the CSIDH moves around the surface. The curve coefficient of the image curve can be recovered by a 2-torsion point. We also proved that the CSIDH using the proposed parameter guarantees a free and transitive group action. Additionally, we present the implementation result using our method. We demonstrated that our method is 6.1% faster than the original CSIDH. Our works show that quite higher performance of CSIDH is achieved using only Montgomery curves.

The Franco-Prussian war (1870--1871) was the first major European conflict during which extensive telegraph use enabled fast communication across large distances. Field officers would therefore have to learn how to use secret codes. But training officers also raises the probability that defectors would reveal these codes to the enemy. Practically all known secret codes at the time could be broken if the enemy knew how they worked. Under Kerckhoffs' impulsion, the French military thus developed new codes, meant to resist even if the adversary knew the encoding and decoding algorithms, but simple enough to be explained and taught to military personnel. Many of these codes were lost to history. One of the designs however, due to Major H. D. Josse, has been recovered and this article describes the features, history, and role of this particular construction. Josse's code was considered for field deployment and underwent some experimental tests in the late 1800s, the result of which were condensed in a short handwritten report. During World War II, German forces got hold of documents describing Josse's work, and brought them to Berlin to be analyzed. A few years later these documents moved to Russia, where they have resided since.

All mainstay ciphers share an underemphasized vulnerability: their ciphertext commits to its generating plaintext. This means that fast enough computers will cryptanalyze them, and so will an attacker smarter than their designers. By contrast, the Vernam One-Time-Pad cipher is free from these vulnerabilities, which is why it is the cipher of choice against such perceived threats. Alas, Vernam key management is very exacting and cumbersome, and it is also plagued by a serious authentication vulnerability. It is therefore of some interest to use a cipher that shares the mathematical secrecy delivered by Vernam, while overcoming its weaknesses. Such is the here proposed unary cipher. It uses the fundamental aspect of Vernam -- a very large key, and takes it even further -- an even larger key space. As a result the unary cipher exhibits good resilience to re-use of the same key (no resilience with Vernam), and it is also immunized to the Vernam authentication flaw. The unary cipher re-writes the plaintext in a unary alphabet, allows it to be mixed with contents-free bits, and then it transposes the resultant plaintext. Since it is possible to build the plaintext out of contents-free bits only, then use of the unary cipher successfully hides usage pattern. It is shown that the transposed message can be reverse-transposed to every plaintext up to a certain size. This plaintext variety is the same principle Vernam relies on to deliver its mathematical secrecy. The unary cipher offers a disadvantage in the form of a larger ciphertext compared to Vernam, and so its practical use will have to be carefully evaluated.

This paper introduces 4 way vectorization of Montgomery ladder on any Montgomery form elliptic curve. Our algorithm takes 2M^4+1S^4 (M^4: A vector of four field multiplications, S^4: A vector of four field squarings) per ladder step for variable-scalar variable-point multiplication. This paper also introduces new formulas for doing arithmetic over GF(2^255-19).

Noisy measurements of a physical unclonable function (PUF) are used to store secret keys with reliability, security, privacy, and complexity constraints. A new set of low-complexity and orthogonal transforms with no multiplication is proposed to obtain bit-error probability results significantly better than all methods previously proposed for key binding with PUFs. The uniqueness and security performance of a transform selected from the proposed set is shown to be close to optimal. An error-correction code with a low-complexity decoder and a high code rate is shown to provide a block-error probability significantly smaller than provided by previously proposed codes with the same or smaller code rates.

While accountability is a well-known concept in distributed systems and cryptography, in the literature on blockchains (and, more generally, distributed ledgers) the formal treatment of accountability has been a blind spot: there does not exist a formalization let alone a formal proof of accountability for any blockchain yet. Therefore, in this work we put forward and propose a formal treatment of accountability in this domain. Our goal is to formally state and prove that if in a run of a blockchain a central security property, such as consistency, is not satisfied, then misbehaving parties can be identified and held accountable. Accountability is particularly useful for permissioned blockchains where all parties know each other, and hence, accountability incentivizes all parties to behave honestly. We exemplify our approach for one of the most prominent permissioned blockchains: Hyperledger Fabric in its most common instantiation.

Private intersection-sum with cardinality allows two parties, where each party holds a private set and one of the parties additionally holds a private integer value associated with each element in her set, to jointly compute the cardinality of the intersection of the two sets as well as the sum of the associated integer values for all the elements in the intersection, and nothing beyond that. We present a new construction for private intersection sum with cardinality that provides malicious security with abort and guarantees that both parties receive the output upon successful completion of the protocol. A central building block for our constructions is a primitive called shuffled distributed oblivious PRF (DOPRF), which is a PRF that offers oblivious evaluation using a secret key shared between two parties, and in addition to this allows obliviously permuting the PRF outputs of several parallel oblivious evaluations. We present the first construction for shuffled DOPRF with malicious security. We further present several new sigma proof protocols for relations across Pedersen commitments, ElGamal encryptions, and Camenisch-Shoup encryptions that we use in our main construction, for which we develop new batching techniques to reduce communication. We implement and evaluate the efficiency of our protocol and show that we can achieve communication cost that is only 4-5 times greater than the most efficient semi-honest protocol. When measuring monetary cost of executing the protocol in the cloud, our protocol is 25 times more expensive than the semi-honest protocol. Our construction also allows for different parameter regimes that enable trade-offs between communication and computation.

This article discusses a fixed critical security bug in Google Tink's Ed25519 Java implementation. The bug allows remote attackers to extract the private key with only two Ed25519 signatures. The vulnerability comes from the misunderstanding of what "final" in Java programming language means. The bug was discovered during security review before Google Tink was officially released. It reinforces the challenge in writing safe cryptographic code and the importance of the security review process even for the code written by professional cryptographers.

We present novel strategies and concrete algorithms for the parallel computation of the Supersingular Isogeny-based Diffie-Hellman key exchange (SIDH) protocol when executed on multi-core platforms. The most relevant design idea exploited by our approach is that of concurrently computing scalar multiplication operations along with a parallelized version of the strategies required for constructing and evaluating large smooth degree isogenies. We report experimental results showing that a three-core implementation of our parallel approach achieves an acceleration factor of 1.56 compared against a sequential implementation of the SIKE protocol.