Search

International audience; The Module Learning With Errors problem (M-LWE) has gained popularity in recent years for its security-efficiency balance, and its hardness has been established for a number of variants. In this paper, we focus on proving the hardness of (search) M-LWE for general secret distributions, provided they carry sufficient min-entropy. This is called entropic hardness of M-LWE. First, we adapt the line of proof of Brakerski and Döttling on R-LWE (TCC'20) to prove that the existence of certain distributions implies the entropic hardness of M-LWE. Then, we provide one such distribution whose required properties rely on the hardness of the decisional Module-NTRU problem.

International audience; Local energy markets allow neighbours to exchange energy among them. Their traditional implementation using sequential auctions has proven to be inefficient and even counterproductive in some cases. In this paper we propose a combinatorial double auction for the exchange of energy for several time-slots simultaneously. We suppose that participants have a flexible demand; flexibility being obtained, for example, by the usage of a battery. We show the benefits of the approach and we provide an example of how it can improve the utility of all the participants in the market.

International audience; One major advance in software security would be to use robust processors which could assist the code developer to thwart both cyber and physical attacks. This paper presents a hardware-based solution which increases the security by checking the integrity of executed code on any microcontroller. Unlike other Control Flow Integrity (CFI) protections, this solution does not require modifications of the CPU pipeline, but relies on monitoring the interface between the processor and its instruction cache. The integrity of the execution flow and the instruction sequences (called Basic Blocks) is checked by hardware with precomputed metadata. Another module is dedicated to speed up the access to these metadata. This paper shows the effectiveness of the solution as the impact is as much as 21% in average on the execution time at the price of using memory space to store metadata along with the code.

International audience; Persistent fault analysis (PFA) consists in guessing block cipher secret keys by biasing their substitution box. This paper improves the original attack of Zhang et al. on AES-128 presented at CHES 2018. By a thorough analysis, the exact probability distribution of the ciphertext (under a uniformly distributed plaintext) is derived, and the maximum likelihood key recovery estimator is computed exactly. Its expression is turned into an attack algorithm, which is shown to be twice more efficient in terms of number of required encryptions than the original attack of Zhang et al. This algorithm is also optimized from a computational complexity standpoint. In addition, our optimal attack is naturally amenable to key enumeration, which expedites full 16-bytes key extraction. Various tradeoffs between data and computational complexities are investigated.

Physically Unclonable Functions (PUFs) are well-known to be solutions for silicon-level anti-copy applications. However, as they are sensitive components, they are the obvious target of physical attacks. Thus, they shall be well protected. In this work we discuss the use case of key generation with a Loop PUF. We discuss the Loop PUF’s efficiency and efficacy. We analyze it with respect to several known attacks like side-channel and machine learning attacks, and show that in all considered cases it either natively resists or can be protected. We also show that perturbation attempts should be within the scope of likely attacks, hence the PUF shall be protected against tampering attacks as well. Also for this attack scenario we highlight the salient features of the Loop PUF and explain how its mode of operation natively empowers it to resist such attacks.

We show how to enhance a classical ballot box to enable automatic tally, while keeping the voter experience as close as possible to the one she already knows. We describe the physical add-on, based on off-the-shelf infrared technology, as well as the cryptographic aspects, for which we rely on self-enforcing e-voting systems by Hao et al.

International audience; We give an informative review of the notions of Rényi's αentropy and α-divergence, Arimoto's conditional α-entropy, and Sibson's α-information, with emphasis on the various relations between them. All these generalize Shannon's classical information measures corresponding to α = 1. We present results on data processing inequalities and provide some new generalizations of the classical Fano's inequality for any α > 0. This enables one to α-information as a information theoretic metric of leakage in secrecy systems. Such metric can bound the gain of an adversary in guessing some secret (any potentially random function of some sensitive dataset) from disclosed measurements, compared with the adversary's prior belief (without access to measurements).

Convolutional Neural Networks (CNNs) are the target of several side-channel attacks aiming at recovering their parameters and hyper-parameters. Attack vectors include monitoring of the cache, power consumption analysis and execution time measurements. These attacks often rely on the knowledge of a certain – large – set of hyper-parameters among which the victim model lies. The goal of the potential attacker is then to reduce that search space or even deduce the correct architecture. One such attack, Cache Telepathy by Yan et al., monitors access to a common matrix multiplication algorithm, GeMM (Generalized Matrix Multiply), in order to determine the victim model’s hyper-parameters. In this paper, we propose to change the order in which the computations are made and add randomness to the said computations in order to mitigate Cache Telepathy. The security analysis of our protection shows that the Cache Telepathy attack on a protected VGG-16 has an increased search space: from 16 to \(2^{22}\).

International audience; This paper aims at providing an overview of the use of mathematical morphology, in its algebraic setting, in several fields of artificial intelligence (AI). Three domains of AI will be covered. In the first domain, mathematical morphology operators will be expressed in some logics (propositional, modal, description logics) to answer typical questions in knowledge representation and reasoning, such as revision, fusion, explanatory relations, satisfying usual postulates. In the second domain, spatial reasoning will benefit from spatial relations modeled using fuzzy sets and morphological operators, with applications in model-based image understanding. In the third domain, interactions between mathematical morphology and deep learning will be detailed. Morphological neural networks were introduced as an alternative to classical architectures, yielding a new geometry in decision surfaces. Deep networks were also trained to learn morphological operators and pipelines, and morphological algorithms were used as companion tools to machine learning, for pre/post processing or even regularization purposes. These ideas have known a large resurgence in the last few years and new ones are emerging.

International audience; Physical Unclonable Functions (PUFs) provide means to gen-erate chip individual keys, especially for low-cost applications such as theInternet of Things (IoT). They are intrinsically robust against reverseengineering, and more cost-effective than non-volatile memory (NVM).For several PUF primitives, countermeasures have been proposed to mit-igate side-channel weaknesses. However, most mitigation techniques re-quire substantial design effort and/or complexity overhead, which can-not be tolerated in low-cost IoT scenarios. In this paper, we first ana-lyze side-channel vulnerabilities of the Loop PUF, an area efficient PUFimplementation with a configurable delay path based on a single ringoscillator (RO). We provide side-channel analysis (SCA) results frompower and electromagnetic measurements. We confirm that oscillationfrequencies are easily observable and distinguishable, breaking the se-curity of unprotected Loop PUF implementations. Second, we presenta low-cost countermeasure based on temporal masking to thwart SCAthat requires only one bit of randomness per PUF response bit. The ran-domness is extracted from the PUF itself creating aself-secured PUF.The concept is highly effective regarding security, low complexity, andlow design constraints making it ideal for applications like IoT. Finally,we discuss trade-offs of side-channel resistance, reliability, and latency aswell as the transfer of the countermeasure to other RO-based PUFs

Fault attacks have raised serious concern with the growing amount of connected devices. Even a small vulnerability might compromise a complete network. It is therefore important to secure all the devices in the connected architecture. A solution to this problem is presented in this paper where we provide a hardware framework, called Smart Monitor, that utilizes a set of sensors (digital or physical) placed on the chip alongside the security target to be protected. The framework continuously monitors the status of the sensors and its Artificial Intelligence (AI) core produces two outputs viz. presence of a fault and type of detected perturbation. The types of attack sources can be electromagnetic, clock-glitch, laser, temperature, etc. In this work we utilize Electro-Magnetic (EM) and Clock-Glitch (CG) as sources of fault injections. Both attacks are performed in multiple settings to increase attack diversity. The framework is able to detect the presence of an attack with 92% accuracy for mixed or multiple attack sources, and further classify the type of perturbation with 78% accuracy keeping the false positive rate at 0%. Overall, this two-stage detection framework is a cost-effective countermeasure that can be deployed easily in any integrated circuit to safeguard against multiple fault attacks. The AI core is further evaluated for consistency in performance on hardware using High Level Synthesis (HLS), as a proof-of-concept, emulating real-world scenario.

We propose algorithms for construction and random generation of hypergraphs without loops and with prescribed degree and dimension sequences. The objective is to provide a starting point for as well as an alternative to Markov chain Monte Carlo approaches. Our algorithms leverage the transposition of properties and algorithms devised for matrices constituted of zeros and ones with prescribed row- and column-sums to hypergraphs. The construction algorithm extends the applicability of Markov chain Monte Carlo approaches when the initial hypergraph is not provided. The random generation algorithm allows the development of a self-normalised importance sampling estimator for hypergraph properties such as the average clustering coefficient.We prove the correctness of the proposed algorithms. We also prove that the random generation algorithm generates any hypergraph following the prescribed degree and dimension sequences with a non-zero probability. We empirically and comparatively evaluate the effectiveness and efficiency of the random generation algorithm. Experiments show that the random generation algorithm provides stable and accurate estimates of average clustering coefficient, and also demonstrates a better effective sample size in comparison with the Markov chain Monte Carlo approaches., Comment: 21 pages, 3 figures

International audience; Locating-domination and identification are two particular, related, types of domination: a set C of vertices in a graph G = (V, E) is a locating-dominating code if it is dominating and any two vertices of V \ C are dominated by distinct sets of codewords; C is an identifying code if it is dominating and any two vertices of V are dominated by distinct sets of codewords. This chapter presents a survey of the major results on locating-domination and on identification.

International audience; Finding understandable and meaningful feature representation of multivariate time series (MTS) is a difficult task, since information is entangled both in temporal and spatial dimensions. In particular, MTS can be seen as the observation of simultaneous causal interactions between dynamical variables. Standard way to model these interactions is the vector linear autoregression (VAR). The parameters of VAR models can be used as MTS feature representation. Yet, VAR cannot generalize on new samples, hence independent VAR models must be trained to represent different MTS. In this paper, we propose to use the inference capacity of neural networks to overpass this limit. We propose to associate a relational neural network to a VAR generative model to form an encoder-decoder of MTS. The model is denoted Seq2VAR for Sequence-to-VAR. We use recent advances in relational neural network to build our MTS encoder by explicitly modeling interactions between variables of MTS samples. We also propose to leverage reparametrization tricks for binomial sampling in neural networks in order to build a sparse version of Seq2VAR and find back the notion of Granger causality defined in sparse VAR models. We illustrate the interest of our approach through experiments on synthetic datasets.

The complexity of heterogeneous Multi-Processor Systems-on-Chip stretches the limits of software development solutions based on sequential languages such as C/C++. While these are still the most widely used languages in practice, model-based solutions appear to be an efficient alternative. However, the optimized compilation of models for multi-processor systems still presents many open research problems. Among others, static data-flow analyses for models require the adaptation of traditional algorithms used in program analysis (iterative and worklist algorithms). These algorithms operate on Control-Flow Graphs with a unique start node (i.e., a node without predecessors) and assume that every basic block is reachable from this start node.

Private Set Intersection Cardinality (PSI-CA) allows two parties, each holding a set of items, to learn the size of the intersection of those sets without revealing any additional information. To the best of our knowledge, this work presents the first protocol that allows one of the parties to delegate PSI-CA computation to untrusted servers. At the heart of our delegated PSI-CA protocol is a new oblivious distributed key PRF (Odk-PRF) abstraction, which may be of independent interest.

International audience; This article gives an overview on knowledge harvesting: automatically constructing large high-quality knowledge bases from Internet sources. The first part reviews key principles and best-practice methods. The second part points out open challenges for future research.

We obtain concrete upper bounds on the algebraic immunity of a class of highly nonlinear plateaued functions without linear structures than the one was given recently in 2017, Cusick. Moreover, we extend Cusick’s class to a much bigger explicit class and we show that our class has better algebraic immunity by an explicit example. We also give a new notion of linear translator, which includes the Frobenius linear translator given in 2018, Cepak, Pasalic and Muratovic-Ribic as a special case. We find some applications of our new notion of linear translator to the construction of permutation polynomials. Furthermore, we give explicit classes of permutation polynomials over \(\mathbb {F}_{q^n}\) using some properties of \(\mathbb {F}_q\) and some conditions of 2011, Akbary, Ghioca and Wang.

International audience; Data provenance is extra information computed during query evaluation over databases, which provides additional context about query results. Several formal frameworks for data provenance have been proposed , in particular based on provenance semirings. The provenance of a query can be computed in these frameworks for a variety of query languages. Provenance has applications in various settings, such as prob-abilistic databases, view maintenance, or explanation of query results. Though the theory of provenance semirings has mostly been developed in the setting of relational databases, it can also apply to other data representations, such as XML, graph, and triple-store databases.

We present a research trajectory of the authors and colleagues dealing with the correctness and meaningful composition of software components, trajectory that incrementally traverses successive paradigms and approaches: open distributed processing, contract based reasoning, behavioural typing and session types. This research is grounded on the foundational work of Robin Milner on processes and observation equivalence, and the followup work by De Nicola and Hennessy on testing relations. Indeed, these initial works have set benchmarks that define the meaning of behaviour, which has fostered a full body of research in concurrency and verification. Behavioural typing is one of the avenues opened by these early contributions. This paper is a brief and staged report of the research accomplished by the authors and colleagues, presented in chronological order, starting with their work on the computational model of open distributed processing and ending at their latest work on sessions for web services.

Numerous masking schemes have been designed as provable countermeasures against side-channel attacks. However, currently, several side-channel attack models coexist, such as “probing” and “bounded moment” models, at bit or word levels. From a defensive standpoint, it is thus unclear which protection strategy is the most relevant to adopt.

International audience; Originally used to assist network-layer fragmentation and reassembly, the IP identification field (IP-ID) has been used and abused for a range of tasks, from counting hosts behind NAT, to detect router aliases and, lately, to assist detection of censorship in the Internet at large. These inferences have been possible since, in the past, the IP- ID was mostly implemented as a simple packet counter: however, this behavior has been discouraged for security reasons and other policies, such as random values, have been suggested. In this study, we propose a framework to classify the different IP-ID behaviors using active probing from a single host. Despite being only minimally intrusive, our technique is significantly accurate (99% true positive classification) robust against packet losses (up to 20%) and lightweight (few packets suffices to discriminate all IP-ID behaviors). We then apply our technique to an Internet-wide census, where we actively probe one alive target per each routable /24 subnet: we find that that the majority of hosts adopts a constant IP-IDs (39%) or local counter (34%), that the fraction of global counters (18%) significantly diminished, that a non marginal number of hosts have an odd behavior (7%) and that random IP-IDs are still an exception (2%).

In this paper, first given a directed network G and a vertex \(r \in V(G)\), we propose a new exact algorithm to compute betweenness score of r. Our algorithm pre-computes a set \(\mathcal {RF}(r)\), which is used to prune a huge amount of computations that do not contribute in the betweenness score of r. Then, for the cases where \(\mathcal {RF}(r)\) is large, we present a randomized algorithm that samples from \(\mathcal {RF}(r)\) and performs computations for only the sampled elements. We show that this algorithm provides an \((\epsilon ,\delta )\)-approximation of the betweenness score of r. Finally, we empirically evaluate our algorithms and show that they significantly outperform the most efficient existing algorithms, in terms of both running time and accuracy. Our experiments also show that our proposed algorithms can effectively compute betweenness scores of all vertices in a set of vertices.

International audience; Most of the time it is nearly impossible to differentiate between particular type of sound events from a waveform only. Therefore, frequency domain and time-frequency domain representations have been used for years providing representations of the sound signals that are more inline with the human perception. However, these representations are usually too generic and often fail to describe specific content that is present in a sound recording. A lot of work have been devoted to design features that could allow extracting such specific information leading to a wide variety of hand-crafted features. During the past years, owing to the increasing availability of medium scale and large scale sound datasets, an alternative approach to feature extraction has become popular, the so-called feature learning. Finally, processing the amount of data that is at hand nowadays can quickly become overwhelming. It is therefore of paramount importance to be able to reduce the size of the dataset in the feature space. The general processing chain to convert an sound signal to a feature vector that can be efficiently exploited by a classifier and the relation to features used for speech and music processing are described is this chapter.

Part 7: Pointing and Target Selection; International audience; Movement time in Fitts’ law is usually considered through the ambiguous notion of the average of minimum movement times. In this paper, we argue that using two distinct metrics, one relating to minimum time and the other relating to average time can be advantageous. Both metrics have a lot of support from theoretical and empirical perspectives. We also give two examples, one in a controlled experiment and the other in a field study of pointing, where making the minimum versus average distinction is fruitful.

International audience; We consider the base stations location problem with sharing. New operator arrives the market and competes with an existing network of another operator. Latter one can share BS cites with new operator, receiving a rent payment from him. We propose new model of realistic clients behavior and formulate the problem as a nonlinear integer programming problem. We propose a fast tabu search heuristic for this problem and provide some computational results.

We define the notion of “self-aware computing” and the relationship of this term to related terms such as autonomic computing, self-management, and similar. The need for a new definition, driven by trends that are only partially addressed by existing areas of research, is motivated. The semantics of the provided definition are discussed in detail examining the selected wording and explaining its meaning to avoid misleading interpretations. This chapter also provides an overview of the existing usage of the term self-aware computing, respectively self-awareness, in related past projects and initiatives.

The collection and exploitation of ratings from users are modern pillars of collaborative filtering. Likert scale is a psychometric quantifier of ratings popular among the electronic commerce sites. In this paper, we consider the tasks of collecting Likert scale ratings of items and of finding the n-k best-rated items, i.e., the n items that are most likely to be the top-k in a ranking constructed from these ratings. We devise an algorithm, Pundit, that computes the n-k best-rated items. Pundit uses the probability-generating function constructed from the Likert scale responses to avoid the combinatorial exploration of the possible outcomes and to compute the result efficiently. Selection of the best-rated items meets, in practice, the major obstacle of the scarcity of ratings. We propose an approach that learns from the available data how many ratings are enough to meet a prescribed error. We empirically validate with real datasets the effectiveness of our method to recommend the collection of additional ratings.

International audience; In many domains of information processing,both vagueness, or imprecision, and bipolarity, encompassing positive and negative parts of information, are core features of the information to be modeled and processed. This led to the development of the concept of bipolar fuzzy sets, and of associated models and tools.Here we propose to extend these tools by defining set theoretical and topological relations between bipolar fuzzy sets, including intersection, inclusion, adjacency and RCC relations widely used in mereotopology,based on bipolar connectives and on mathematical morphology operators.

TLS is currently receiving much attention and seems to be the most promising candidate as a secure protocol for the current and future online services. Many security protocols have been developed to handle security problems in various circumstances. However, these protocols are often limited and related to precised contexts and architectures and cannot answer special application and networks needs such as securing transactions between several vehicles in vehicular networks. In this paper, we propose to extend the Transport Layer Security (TLS) protocol by using ETSI and IEEE certificates to securely exchange data between multiple vehicular network components through a single TLS session. We propose a formal validation of our extension with the HLPSL language using AVISPA. The proposed extension is currently being implemented over OPENSSL.

The collection and exploitation of ratings from users are modern pillars of collaborative filtering. Likert scale is a psychometric quantifier of ratings popular among the electronic commerce sites. In this paper, we consider the tasks of collecting Likert scale ratings of items and of finding the n-k best-rated items, i.e., the n items that are most likely to be the top-k in a ranking constructed from these ratings. We devise an algorithm, Pundit, that computes the n-k best-rated items. Pundit uses the probability-generating function constructed from the Likert scale responses to avoid the combinatorial exploration of the possible outcomes and to compute the result efficiently. Selection of the best-rated items meets, in practice, the major obstacle of the scarcity of ratings. We propose an approach that learns from the available data how many ratings are enough to meet a prescribed error. We empirically validate with real datasets the effectiveness of our method to recommend the collection of additional ratings.

International audience; The book collects contributions from experts worldwide addressing recent scholarship in social network analysis such as influence spread, link prediction, dynamic network biclustering, and delurking. It covers both new topics and new solutions to known problems. The contributions rely on established methods and techniques in graph theory, machine learning, stochastic modelling, user behavior analysis and natural language processing, just to name a few. This text provides an understanding of using such methods and techniques in order to manage practical problems and situations. Trends in Social Network Analysis: Information Propagation, User Behavior Modelling, Forecasting, and Vulnerability Assessment appeals to students, researchers, and professionals working in the field.

International audience; A new generation of Autonomous systems (UAVs, ROVERs, etc.) is coming that will help improve the situational awareness and assessment, especially in difficult conditions like disasters. Rescuers should be relieved from time-consuming data collection tasks as much as possible and at the same time, Autonomous systems should assist data collection through a more insightful and automated guidance thanks to advanced sensing capabilities. In order to achieve this vision, two challenges must be addressed though. The first one is to achieve a sufficient autonomy. The second one relates to the reliability with respect to accidental (safety) or even malicious (security) risks. This however requires the design of new embedded architectures to be more autonomous, while mitigating the harm they may potentially cause. Increased complexity and flexibility requires resorting to modelling, simulation and formal verification techniques in order to validate such critical aspects.

International audience; One of the major gaps in the current HTML5 web platform is the lack of an interoperable means for a multimodal application to discover services and applications available in a given space and network, for example, in a smart house with a network of connected objects. To address this gap, the Multimodal Interaction Working Group has produced a draft specification based on distributed services, which aims to support the Discovery and Registration of multimodal components. In this approach, the components are described and virtualized in a Resources Manager communicating bidirectionally through dedicated events. To facilitate the fine-grained management of concurrent multimodal interactions, the Resources Manager registers the distributed components and provides to the Interaction Manager the means to control them. In this way, interoperable search, discovery, and selection of heterogeneous and dynamic features on the Web of Things can be performed by multimodal applications producing natural interaction and a semantically rich user experience.

A failure detector is a distributed oracle that provides each process with a module that continuously outputs an estimate of which processes in the system have failed. The perfect failure detector provides accurate and eventually complete information about process failures. We show that, in asynchronous failure-prone message-passing systems, perfect failure detection can be achieved by an oracle that outputs at most \(\lceil \log \alpha (n)\rceil +1\) bits per process in n-process systems, where \(\alpha \) denotes the inverse-Ackermann function. This result is essentially optimal, as we also show that, in the same environment, no failure detector outputting a constant number of bits per process can achieve perfect failure detection.

Developing complex safety-critical systems usually involves developing models as abstractions in the upstream phases of design. It is still today often challenging to convince the industry that performing functional testing on models of systems may help reducing the cost of system testing. This article presents a new model-based testing process. Part of the “CONNEXION” French I&C methodology project, it combines a vast number of verification tools. In this article, we focus on the integration of a specification-based test generation tool, a model-checker and an environment for model test execution to enhance structural coverage rate. To this end, we define a novel process describing how to extend the functional test bed to enhance structural coverage by generating new test cases reaching so far uncovered branches using model-checking.

International audience; In this paper, we propose a learning approach to adaptive performance tuning of database applications. The objective is to validate the opportunity to devise a tuning strategy that does not need prior knowledge of a cost model. Instead, the cost model is learned through reinforcement learning. We instantiate our approach to the use case of index tuning. We model the execution of queries and updates as a Markov decision process whose states are database configurations, actions are configuration changes, and rewards are functions of the cost of configuration change and query and update evaluation. During the reinforcement learning process, we face two important challenges: not only the unavailability of a cost model, but also the size of the state space. To address the latter, we devise strategies to prune the state space, both in the general case and for the use case of index tuning. We empirically and comparatively evaluate our approach on a standard OLTP dataset. We show that our approach is competitive with state-of-the-art adaptive index tuning, which is dependent on a cost model.

International audience; Generic web crawling approaches cannot distinguish among various page types and cannot target content-rich areas of a website. We study the problem of efficient unsupervised web crawling of content-rich webpages. We propose ACEBot (Adaptive Crawler Bot for data Extraction), a structure-driven crawler that uses the inner structure of the pages and guides the crawling process based on the importance of their content. ACEBot works in two phases: in the learning phase, it constructs a dynamic site map (limiting the number of URLs retrieved) and learns a traversal strategy based on the importance of navigation patterns (selecting those leading to valuable content); in the intensive crawling phase, ACEBot performs massive downloading following the chosen navigation patterns. Experiments over a large dataset illustrate the effectiveness of our system.

International audience; With the aim of estimating the growth of tomatoes duringthe agricultural season, we propose to segment tomatoes in imagesacquired in open field, and to derive their size from the segmentationresults obtained in pairs of images acquired each day. To cope withdifficult conditions such as occlusion, poor contrast and movement oftomatoes and leaves, we propose to base the segmentation of an imageon the result obtained on the image of the previous day, guaranteeingtemporal consistency, and to incorporate a shape constraint in the segmentationprocedure, assuming that the image of a tomato is approximatelyan ellipse, guaranteeing spatial consistency. This is achieved witha parametric deformable model with shape constraint. Results obtainedover three agricultural seasons are very good for images with limitedocclusion, with an average relative distance between the automatic andmanual segmentations of 6.46% (expressed as percentage of the size oftomato).

The characteristic independence property of Poisson point processes gives an intuitive way to explain why a sequence of point processes becoming less and less repulsive can converge to a Poisson point process. The aim of this paper is to show this convergence for sequences built by superposing, thinning or rescaling determinantal processes. We use Papangelou intensities and Stein’s method to prove this result with a topology based on total variation distance.

The introduction of security flaws into a system may result from design or implementation mistakes. It entail far-reaching consequences for connected embedded or cyber-physical systems, including physical harm. Security experts focus either on finding out and deriving security mechanisms from more or less explicitly defined security requirements or on the a posteriori assessment of vulnerabilities, namely pentesting. These approaches however often miss the necessary iterations between security countermeasures and system functionalities in terms of design and deployment. Worse, they generally fail to consider the implications of security issues over the system’s safety, like for instance the adverse effect that security countermeasures may produce over expected deadlines due to costly computations and communications latencies. SysML-Sec focuses on these issues throughout design and development thanks to its model-driven approach that promotes exchanges between system architects, safety engineers, and security experts. This paper discusses how SysML-Sec can be used to simultaneously deal with safety and security requirements, and illustrates the methodology with an automotive use case.

This chapter presents a novel Physically Unclonable Function PUF called “Loop PUF” (LPUF) which has been studied specifically to be easy to design, lightweight and reliable. The LPUF principle is based on a loop of N controllable delay lines forming a unique ring oscillator. It offers a huge set of challenges as the identity extraction is performed by N measurements, whereas the extraction is made differentially for the other delay-based PUF. This property allows the designer to forge stronger challenge-response protocols or to generate more reliable internal keys. The LPUF concept has been designed and evaluated in ASIC 65 nm technology and in FPGA Virtex-5. The evaluation results show good properties of randomness and uniqueness. As the LPUF output is in integer format and can use numerous challenges, it provides a good base to enhance the steadiness and build reliable authentication protocols. This chapter presents an example of authentication primitive which is very steady in a large range of environmental conditions.

Multi-variate side-channel attacks allow to break higher- order masking protections by combining several leakage samples. But how to optimally extract all the information contained in all possible d-tuples of points? In this article, we introduce preprocessing tools that answer this question. We first show that maximizing the higher-order CPA coefficient is equivalent to finding the maximum of the covariance. We apply this equivalence to the problem of trace dimensionality reduction by linear combination of its samples. Then we establish the link between this problem and the Principal Component Analysis. In a second step we present the optimal solution for the problem of maximizing the covariance. We also theoretically and empirically compare these methods. We finally apply them on real measurements, publicly available under the DPA Contest v4, to evaluate how the proposed techniques improve the second-order CPA (2O-CPA).

International audience; In this paper, we carry out a detailed mathematical study of two theoretical distinguishers based on the Kolmogorov-Smirnov (KS) distance. This includes a proof of soundness and the derivation of closed-form expressions, which can be split into two factors: one depending only on the noise and the other on the confusion coefficient of Fei, Luo and Ding. This allows one to have a deeper understanding of the relative influences of the signal-to-noise ratio and the confusion coefficient on the distinguisher’s performance. Moreover, one is able to directly compare distinguishers based on their closed-form expressions instead of using evaluation metric that might obscure the actual performance and favor one distinguisher over the other. Furthermore, we formalize the link between the confusion coefficient and differential cryptanalysis, which shows that the stronger an S-box is resistant to differential attacks the weaker it is against side-channel attacks, and vice versa.

Side-channel attacks are considered as one of the biggest threats against modern crypto-systems. This motivates the design of ciphers which are naturally resistant against side-channel attacks. The present paper proposes a scheme called DRECON to construct a block cipher with innate protection against differential power attacks (DPA). The scheme is motivated by tweakable block ciphers and is shown to be secure against first-order DPA using information theoretic metrics. DRECON is shown to be less expensive than masking and re-keying countermeasures from the implementation perspective and can be efficiently realized in both hardware and software platforms. On FPGAs especially, DRECON can optimally utilize the abundant block RAMs available and therefore have minimal overheads. We estimate the cost overhead of DRECON in micro-controllers and FPGAs, two common targets for cryptographic applications. Finally we demonstrate practical side-channel resistance of a DRECON implementation on a Xilinx Virtex-5 FPGA (SASEBO GII board).

International audience; The SAIBA framework proposes two interface languages to representseparately an intelligent agent’s communicative functions (or intents) andthe multimodal behavior determining how the functions are accomplished with aparticular multimodal realization. For the functional level, the Function MarkupLanguage (FML) has been proposed. In this paper we summarize the current statusof FML as discussed by the SAIBA community, we underline the major issuesthat need to be addressed to obtain a unified FML specification, we suggest furtherissues that we identified and we propose a new unified FML specificationthat addresses many of these issues.

This chapter provides an overview of how deep learning techniques can be used for audio signals. We first review the main DNN architectures, meta-architectures and training paradigms used for audio processing. By highlighting the specifies of the audio signal, we discuss the various possible audio representations to be used as input of a DNN—time and frequency representations, waveform representations and knowledge-driven representations—and discuss how the first layers of a DNN can be set to take into account these specificity’s. We then review a set of applications for three main classes of problems: audio recognition, audio processing and audio generation. We do this considering two types of audio content which are less commonly addressed in the literature: music and environmental sounds.

Embedded systems are commonly built upon heterogeneous digital and analog integrated circuits, including sensors and actuators. Model-driven approaches for designing software and hardware are generally limited to the digital parts of systems. In the present paper, we adopt a global view on the extensions made to an integrated modeling and simulation tool, TTool. In this tool, the verification and virtual prototyping of embedded systems is described at different abstraction levels and extended in order to handle analog/mixed-signal systems. An extensive case study spans these levels and illustrates the usefulness of our approach.

International audience; The methods for acquiring, processing, and visualizing magnetoencephalography (MEG) and electroencephalography (EEG) data are rapidly evolving. Advancements in hardware and software development offer new opportunities for cognitive and clinical neuroscientists but at the same time introduce new challenges as well. In recent years the MEG/EEG community has developed a variety of software tools to overcome these challenges and cater to individual research needs. As part of this endeavor, the MNE software project, which includes MNE-C, MNE-Python, MNE-CPP, and MNE-MATLAB as its subprojects, offers an efficient set of tools addressing certain common needs. Even more importantly, the MNE software family covers diverse use case scenarios. Here, we present the landscape of the MNE project and discuss how it will evolve to address the current and emerging needs of the MEG/EEG community.