Publications

2020

• Impact of Security Measures on Performance Aspects in SysML Models
  
  • Zoor Maysam
  • Apvrille Ludovic
  • Pacalet Renaud
  , 2020, pp.373-380. (10.5220/0008970203730380)
  DOI : 10.5220/0008970203730380
• CoRA: a scalable Collective Remote Attestation protocol for sensor networks
  
  • Diop Aida
  • Laurent Maryline
  • Leneutre Jean
  • Traoré Jacques
  , 2020, pp.84-95. Embedded Internet of Things (IoT) devices are deployed in the functioning of a number of applications such as industrial control, building automation, and the smart grid. The lack of robustness of IoT devices has however rendered such systems vulnerable to a number of remote cyber-attacks. Remote attestation is a security mechanism which enables to remotely verify the integrity of the software running on IoT devices. Similarly, collective remote attestation protocols are designed to efficiently verify the integrity of a group of devices. Existing collective attestation protocols do not provide an efficient and secure mechanism to detect compromised devices. In particular, it is not possible to efficiently trace the origin of an erroneous attestation response back to the concerned node. In this paper, we introduce CoRA, a highly scalable collective attestation protocol, which leverages the aggregating property of the underlying cryptographic scheme during the attestation process. CoRA is the first collective attestation protocol to also provide sequential detection, where the identity of the compromised node is revealed. We provide rigorous security proofs for our protocol and its underlying cryptographic primitive, and demonstrate its efficiency in highly scalable networks. (10.5220/0008962700840095)
  DOI : 10.5220/0008962700840095
• Neural Network Security: Hiding {CNN} Parameters with Guided Grad-CAM
  
  • Guiga Linda
  • Roscoe A. W.
  , 2020.
• High-level Partitioning and Design Space Exploration for Cyber Physical Systems
  
  • Genius Daniela
  • Bournias Ilias
  • Apvrille Ludovic
  • Chotin Roselyne
  , 2020, pp.84-91. Virtual prototyping and co-simulation of mixed analog/digital embedded systems have emerged as a promising research topic, but usually assume an already Hardware/Software partitioned system. The paper presents a new approach for high-level system partitioning of such mixed systems, by expressing the structure and the behaivour of the analog parts with SysML diagrams. A tool already able to handle some aspects of analog design after partitioning has been extended to be able to handle partitioning, thus completing the methodology. As a real-world case study, we show the design of the hardware part of a medical application. (10.5220/0009171600840091)
  DOI : 10.5220/0009171600840091
• Model-Based Virtual Prototyping of CPS: Application to Bio-Medical Devices
  
  • Genius Daniela
  • Bournias Ilias
  • Apvrille Ludovic
  • Chotin Roselyne
  , 2021, 1361, pp.74-96. Virtual prototyping and co-simulation of mixed analog/ digital embedded systems have emerged as a promising research topic, in particular for designing medical appliances. In the paper, we show how the integration of different, analog and digital, Models of Computation (MoC) within an UML/SysML based environment, can offer an efficient assistance for designing a cyber-physical system in a progressive and systematic manner. For this, we rely on formal verification and abstract simulation on a high abstraction level, and on Multi-MoC virtual prototyping on a lower abstraction level. A realistic echo monitoring system illustrates (i) the method, (ii) the modeling languages, and (iii) the different verification techniques. (10.1007/978-3-030-67445-8_4)
  DOI : 10.1007/978-3-030-67445-8_4
• A Reputation System Using a Bayesian Statistical Filter in Vehicular Networks
  
  • Begriche Youcef
  • Khatoun Rida
  • Rachini Ali
  • Khoukhi Lyes
  , 2020, pp.1-7. The rapid development of wireless technologies has led to the emergence of new types of highly dynamic networks with centralized and decentralized architectures. Vehicular networks (VANETs) technology is one of those networks. The communication between vehicles will lead to more efficient and secured roads by providing information about traffic and road conditions to vehicle drivers. In order to protect vehicle passengers and drivers, these networks need a secure infrastructure to prevent malicious behavior from happening. This paper proposes a strong reputation system based on a Bayesian statistical filter to detect malicious nodes which drop packets selectively or completely in VANETs. The proposed approach allows vehicles to interact in order to detect the malicious one, give it a bad reputation and exclude it from the network. (10.1109/MobiSecServ48690.2020.9042960)
  DOI : 10.1109/MobiSecServ48690.2020.9042960
• Time Stamped Bijective MAC and Dynamic PUF Authentication New Directions For IoT Security
  
  • Urien Pascal
  , 2020, pp.1-6. (10.1109/MobiSecServ48690.2020.9042939)
  DOI : 10.1109/MobiSecServ48690.2020.9042939
• Distributed online Data Anomaly Detection for connected vehicles
  
  • Negi Naman
  • Jelassi Ons
  • Chaouchi Hakima
  • Clémençon Stéphan
  , 2020, pp.494-500. Wireless connectivity evolution increased the volume of acquired available data in different Internet of Things based industries. Data quality and processing time are the most challenging issues for successful data analytic algorithms to produce efficient business intelligence. In this article we tackle these two points and propose a distributed framework for data anomaly detection. In fact one of the major issues in systems that depend highly on data is detection of anomalies. Long Short Term Memory (LSTM) based anomaly detection in time series data has been studied in the past with promising results. In this article, we use LSTM model and apply distributed learning approach to train the model. Indeed, using a single machine centralized approach for model training and anomaly detection is not a feasible option when dealing with big amounts of data. Distributed approach improves the training and prediction time, model performance, and allows handle of bigger datasets and higher level of model complexity. We propose a distributed anomaly detection system framework for autonomous and connected cars with a novel online new data selection algorithm that guides the retraining and adjusts the model parameters accordingly. The framework includes the offline training of the LSTM model over many machines in a distributed fashion using all the available data. The trained parameters are then sent to the individual vehicles and the anomaly detection happens at the vehicle level. Finally, the proposed distributed framework is evaluated using MXnet framework, and it shows that with optimized settings we can reduce the model training time, use a more complex LSTM anomaly detection model and improve anomaly detection accuracy. (10.1109/ICAIIC48513.2020.9065280)
  DOI : 10.1109/ICAIIC48513.2020.9065280
• Control Channel Anti-Jamming in Vehicular Networks via Cooperative Relay Beamforming
  
  • Gu Pengwenlong
  • Hua Cunqing
  • Xu Wenchao
  • Khatoun Rida
  • Wu Yue
  • Serhrouchni Ahmed
  IEEE Internet of Things Journal, IEEE, 2020, pp.1-1. (10.1109/JIOT.2020.2973753)
  DOI : 10.1109/JIOT.2020.2973753
• A Combined Stochastic and Physical Framework for Modeling Indoor 5G Millimeter Wave Propagation
  
  • Nassif Georges
  • Gloaguen Catherine
  • Martins Philippe
  , 2020. Indoor coverage is a major challenge for 5G millimeter waves (mmWaves). In this paper, we address this problem through a novel theoretical framework that combines stochastic indoor environment model-ing with advanced physical propagation simulation. This approach is particularly adapted to investigate indoor-to-indoor 5G mmWave propagation. Its system implementation, so-called iGeoStat, generates parameterized typical environments that account for the indoor spatial variations, then simulates radio propagation based on the physical interaction between electromagnetic waves and material properties. This framework is not dedicated to a particular environment, material, frequency or use case and aims to statistically understand the influence of indoor environment parameters on mmWave propagation properties, especially coverage and path loss. Its implementation raises numerous computational challenges that we solve by formulating an adapted link budget and designing new memory optimization algorithms. The first simulation results for two major 5G applications are validated with measurement data and show the efficiency of iGeoStat to simulate multiple diffusion in realistic environments, within a reasonable amount of time and memory resources. Generated output maps confirm that diffusion has a critical impact on indoor mmWave propagation and that proper physical modeling is of the utmost importance to generate relevant propagation models.
• Efficient compression for scalable transmission of digital holograms
  
  • El Rhammad Anas
  , 2020. Contrary to conventional stereoscopy, holography provides the most natural and comfortable 3D visualization. However, digital holograms contain massive amount of data with very few correlations. In the first part of this work, we introduced two methods for digital holograms compression. First, we proposed a redundant light beams-based decomposition of holograms using the Gabor wavelets. For compression purposes, we sparsified the obtained expansion using the Matching Pursuit algorithm. Then, we designed a specific encoder framework for the coefficients and indexes of Gabor atoms. The proposed approach achieved better compression performance compared to the state of the art methods. Second, by exploiting the duality between Gabor wavelets and diffracted light beams, we developed a viewpoint-quality scalable coding scheme. Indeed, for a given observer's position, only the Gabor atoms that emit light into the viewer's window are selected, sorted and then encoded. The bit rate has been significantly reduced, without degrading the reconstruction quality obtained by encoding the whole hologram. In the second part of this work, we designed two server-client architectures for a view-dependent progressive transmission of holograms using scalable coding. In the first solution, a fine-grain scalable bitstream is generated online by the server, after each client notification about the user's position. Experimental results reveal that this method enables a rapid visualization by decoding the first received atoms in addition to a progressive increase of quality.Finally, to reduce the latency caused by the computational burden of encoding, we proposed a second solution where the whole Gabor expansion is encoded offline by the server, and then decoded online with respect to the viewer’s trajectory. To enable a scalable compression, we grouped the Gabor atoms following a block-based decomposition of the observer plane. Then, the atoms of each block are assigned to different quality levels and encoded in packets. Simulations tests show that the proposed architecture allows a low-latency transmission without significantly increasing the encoding rate.
• Electromagnetic aspects of ESPAR and digitally controllable scatterers with a look at low-complexity algorithm design
  
  • Bucheli Garcia Juan
  , 2020. The thesis focuses on the idea of exploiting the spatial domain (as opposed to the exploitation of the time-frequency resource) of wireless environments from two fronts: a) ESPAR antennas (standing for Electronically Steerable Parasitic Array Radiator) as a potential inexpensive alternative to conventional multi-antenna architectures (inexpensive in relation to the number of radio frequency front-ends these conventional architectures are often assumed to be provided with), and b) the study of reactively loaded arrays to deliver controllable scattering as a mean of adding degrees of freedom to the propagation environment itself. The latter is achieved here via digitally controllable scatterers (DCS).Particularly, the thesis focuses on the goal of better conditioning optimization problems as means of proposing low-complexity algorithms. Therefore, one key aspect is the required balance between the accuracy and complexity of the adopted electromagnetic models.Thus, it is appropriate to highlight the importance given to the interface between electromagnetism and the signal characterization. More specifically, both ESPAR and DCS require the understanding of electromagnetic (EM) phenomena that is not fully accounted for through conventional link-level descriptions. More importantly, the latter is proof of the need to join the approaches of two related research communities to cope with the scarcity of resources that is only expected to grow in the decades to come. In fact, the document is mostly positioned from the view of someone with a background in telecommunications (unlike pure electromagnetism) with looks at enlightening the underlying EM mechanisms. It is roughly composed of three parts, namely: fundamentals, the ESPAR antenna and digitally controllable scatterers. In fact, the aim of having one part of the document dedicated purely to fundamentals is to describe the EM phenomena while highlighting all relevant details to the remaining two.The part "fundamentals" begins with Maxwell's equations (and their convenient solution for far-field radiation problems) all the way to the well-known (y=hx+n) signal characterization. As the seemingly least appropriate description to work with, but most complete characterization of the EM phenomena, Maxwell's equations are the basis that link our mathematical description to the very same reality. Thus, the objective of this part is to expose the connection between fields and signals, as well as to open the door to questioning the conventional transmitter-receiver signal model. The latter becomes one of the most exciting outcomes of this project in line with the research-related aim of challenging our vision to expand our understanding of a problem.To continue, the second part is dedicated to the ESPAR antenna as a preamble of what is meant by "questioning the conventional transmitter-receiver signal model". Particularly, ESPAR obliges us to depart from the abstract signal space in which traditional multi-antenna link-level characterizations are depicted. As a contribution of this work, it will be shown how a local approximation of the system model offers an alternative view. Notably, through such an approximation of the system model, a computationally-efficient solution to the non-trivial problem of channel-based adaptation of the radiation characteristics of ESPAR is found.Last, but not least, the third part deals with digitally controllable scatterers as a mean of improving energy efficiency. Such an exciting concept has gained significant attention in the recent years and, in a sense, opens the door to a radically different way to conceive communication problems. Even though these devices are in their infancy, it is not difficult for me to imagine how the decades to come could be marked by the massification of this technology.
• Anthropomorphic devices for affective touch communication
  
  • Teyssier Marc
  , 2020. Communicating emotions is important for human attachment and bonding as well as for physical and psychological well-being.We communicate emotions through voice, but also through body language such as facial expressions, posture or touch. Among all these nonverbal cues, the tactile modality plays a particular role. Touch happens in co-located situations and involves physical contact between two individuals. A touch contact can convey emotions such as comforting someone by gently stroking her arm.Current technologies and devices used for mediated communication are not designed to support affective touch communication.There is a need to have new interfaces to mediate touch, both to detect touch (to replace the receiver's skin) and to convey touch (to replace the emitter's hand).My approach takes inspiration from the human body to inform the design of new interfaces. I promote the use of anthropomorphic affordances to design interfaces that benefit from our knowledge of physical interaction with other humans.Anthropomorphic affordances project human functioning and behaviour to the attributes of an object to suggest ways of using it. However, anthropomorphism has received little attention so far in the field of Human-Computer Interaction; its design space is still largely unexplored. In this thesis, I explore this design space and focus on augmenting mobile and robotic devices with tactile capabilities to enhance the conveying of emotions to enrich social communication.This raises two main research problems addressed in this thesis.A first problem is to define the type of device needed to perform touch. Current actuated devices do not produce human-like touch.In the first part of this thesis, I focus on the design and implementation of interfaces capable of producing humanlike touch output.I highlight human touch factors that can be reproduced by an actuated device. I then experimentally evaluate the impact of humanlike device-initiated touch on the perception of emotions.Finally, I built on top of these findings to propose Mobilimb, a small-scale robotic arm that can be connected onto mobile devices and can touch the user.A second problem is to develop interfaces capable of detecting touch input. My approach is to integrate humanlike artificial skin onto existing devices.I propose requirements to replicate the human skin, and a fabrication method for reproducing its visual, tactile and kinaesthetic properties. I then propose an implementation of artificial skin that can be integrated onto existing devices and can sense expressive touch gestures. This interface is then used to explore possible scenarios and applications for mediated touch input.In summary, this thesis contributes to the design and understanding of anthropomorphic devices for affective touch communication. I propose to use anthropomorphic affordances to design interfaces.To address the research questions of this thesis, I built upon human biological characteristics and digital fabrication tools and methods. The devices presented in this thesis propose new technical and empirical contributions around touch detection and touch generation.
• Machine Learning for Future Wireless Communications
  
  • Ghauch Hadi
  • Luo Fa‐long
  , 2020 (1). (10.1002/9781119562306.ch3)
  DOI : 10.1002/9781119562306.ch3
• Alpha-leakage via Fano's inequality for alpha-information
  
  • Rioul Olivier
  , 2020.
• Rényi Entropy Estimation for Secure Silicon Fingerprints
  
  • Schaub Alexander
  • Rioul Olivier
  • Guilley Sylvain
  • Danger Jean-Luc
  • Boutros Joseph J
  , 2020.
• Multiplexing Gains under Mixed-Delay Constraints on Wyner’s Soft-Handoff Model
  
  • Wigger Michèle
  • Shamai Shlomo
  • Nikbakht Homa
  Entropy, MDPI, 2020, 22 (2), pp.182. This paper analyzes the multiplexing gains (MG) achievable over Wyner's soft-handoff model under mixed-delay constraints, that is, when delay-sensitive and delay-tolerant data are simultaneously transmitted over the network. In the considered model, delay-sensitive data cannot participate or profit in any ways from transmitter or receiver cooperation, but delay-tolerant data can. Cooperation for delay-tolerant data takes place over rate-limited links and is limited to a fixed number of cooperation rounds. For the described setup, inner and outer bounds are derived on the set of MG pairs that are simultaneously achievable for delay-sensitive and delay-tolerant data. The bounds are tight in special cases and allow us to obtain the following conclusions. For large cooperation rates, and when both transmitters and receivers can cooperate, it is possible to simultaneously attain maximum MG for delay-sensitive messages and maximum sum MG for all messages. For comparison, in scheduling schemes (also called time-sharing schemes), the largest achievable sum MG decreases linearly with the MG of delay-sensitive messages. A similar linear decrease is proved for any coding scheme, not only for scheduling schemes, if only transmitters or only receivers can cooperate (but not both) and if delay-sensitive messages have moderate MG. In contrast, if the MG of delay-sensitive messages is small, the maximum sum MG can be achieved even with only transmitter or only receiver cooperation. To summarise, when cooperation rates are high and both transmitters and receivers can cooperate or when delay-sensitive messages have small MG, then transmitting delay-sensitive messages causes no penalty on the sum-MG. In other regimes, this penalty increases proportionally to the delay-tolerant MG in the sense that increasing the delay-sensitive MG by ∆ penalises the largest achievable delay-tolerant MG by 2∆ and thus the sum MG by ∆. (10.3390/e22020182)
  DOI : 10.3390/e22020182
• Epitaxial integration of high-performance quantum-dot lasers on silicon
  
  • Norman Justin
  • Liu Songtao
  • Wan Yating
  • Zhang Zeyu
  • Shang Chen
  • Selvidge Jennifer
  • Dumont Mario
  • Kennedy M.
  • Jung Daehwan
  • Duan Jianan
  • Huang Heming
  • Herrick Robert
  • Grillot Frederic
  • Gossard Arthur
  • Bowers John
  , 2020, pp.2. (10.1117/12.2542912)
  DOI : 10.1117/12.2542912
• Broadcast strategies and performance evaluation of IEEE 802.15.4 in wireless body area networks WBAN
  
  • Badreddine Wafa
  • Chaudet Claude
  • Federico Petruzzi
  • Potop-Butucaru Maria
  Ad Hoc Networks, Elsevier, 2020, pp.102006. Wireless Body Area Networks (WBANs) can be considered as an evolution of wireless sensor networks towards wearable and implanted technologies. Radio propagation and mobility are particular in this context, as they are influenced by the characteristics and movement of the human body and by the necessity to keep the transmission power at its minimum to save energy and limit interactions with the wearer. In this paper, we investigate the broadcasting problem in which a node, typically the gateway, tries to send a packet to all other nodes in the network at minimal cost. This problem is not as trivial as it looks and we show through simulation that forwarding strategies coming from the Delay Tolerant Networks world cannot be transposed without adaptation. We enriched the Omnet++ simulator with a WBAN-specific channel model from the literature, and use this model to evaluate 9 classes of broadcasting algorithms, including our own proposals, with respect to their ability to cover the whole network, their completion delay, their cost in terms of transmissions volume and their capability to preserve multiple packets order (i.e. total order broadcast). Our study shows that there is a subtle compromise to find between verbose strategies that achieve good performance at the cost of numerous transmissions, ultimately provoking collisions and more cautious solutions that miss transmission opportunities because of mobility. (10.1016/j.adhoc.2019.102006)
  DOI : 10.1016/j.adhoc.2019.102006
• PyMarket - A simple library for simulating markets in Python
  
  • Kiedanski Diego
  • Kofman Daniel
  • Horta José
  Journal of Open Source Software, Open Journals, 2020, 5 (46), pp.1591. PyMarket is a python library aimed to ease the design, simulation, and comparison of different market mechanisms. Marketplaces have been proposed to solve a diverse array of problems. They are currently used to sell ads online, allocate bandwidth spectrum, exchange energy, etc. PyMarket provides a simple environment to try, simulate, compare, and visualize different market mechanisms, a task that is inherent to the process of market design. This library is not intended for use in the financial domain, where mature tools already exist1 such as (Chiarella & Iori, 2002),(LeBaron, 2001). Instead, it is targeted at the engineering domain in which markets are sometimes used for interfacing the interaction of multi-agent systems. As an example, Local Energy Markets (LEMs) have been proposed to synchronize energy consumption with a surplus of renewable generation. Several mechanisms have been proposed for such markets, from discrete-time double sided auctions to continuous peer to peer trading. This library aims to provide a simple interface for such processes, making results reproducible. In doing so, it exposes a Market interface that accepts bids, runs market clearing algorithms, and produces statistics and plots (Figure 1) from the results. Moreover, an intuitive procedure is provided to implement new market mechanisms and compare them with existing ones. Algorithms implemented in this library have been used by the authors (Horta, Kofman, Menga, & Silva, 2017) (Kiedanski, Kofman, Horta, & Menga, 2019) as well as other researchers in the field (Mengelkamp, Staudt, Garttner, & Weinhardt, 2017). Moreover, the library is a key enabler of ongoing research in the LEMs. (10.21105/joss.01591)
  DOI : 10.21105/joss.01591
• Towards private optical communications with mid infrared chaotic light
  
  • Grillot F
  • Spitz O
  • Herdt A
  • Elsässer W
  • Carras M
  , 2020, 11288. Free-space optics constitutes a growing technology offering higher bandwidth with fast and cost-effective deployment compared to fiber technology. Multiple applications are envisioned like private communications. In such a case, the secret message is encoded into a chaotic waveform from which the information is extremely hard for an eavesdropper to extract. For free-space optics applications, the operating wavelength is an important parameter that has to be chosen wisely to reduce the impact of the environmental parameters. In this context, quantum cascade lasers are highly relevant semiconductor lasers because the lasing wavelength can be properly adjusted in the mid-infrared domain, typically at wavelengths for which the atmosphere is highly transparent. The simplest way to generate a chaotic optical carrier from a quantum cascade laser is to feed back part of its emitted light into the device after a certain time delay, beyond which chaos synchronization between the drive and the response lasers occurs. In this paper, we discuss about how quantum cascade laser's chaos can be used to develop private communication lines. We also give realistic perspectives for further developing mid-infrared private communications using chaotic waves. (10.1117/12.2546582)
  DOI : 10.1117/12.2546582
• Peculiarities and predictions of rogue waves in mid-infrared quantum cascade lasers under conventional optical feedback
  
  • Spitz O
  • Herdt A
  • Wu J
  • Maisons G
  • Carras M
  • Wong C.-W
  • Elsässer W
  • Grillot F
  , 2020, 11288. Quantum cascade lasers (QCLs) are optical sources exploiting radiative intersubband transitions within the conduction band of semiconductor heterostructures. Mid-infrared QCLs have been thoroughly considered for applications such as spectroscopy, free-space communications and countermeasure systems. Under conventional optical feedback, QCLs have been proven to operate in several non-linear dynamic regimes, including deterministic chaos, entrainment of low-frequency fluctuations and square wave all-optical modulation. We extend the understanding of non-linear phenomena in QCLs with the experimental study of rogue waves. Rogue waves represent random isolated events with amplitudes well above that of neighboring ones, occurring more often than expected from the distribution of lower amplitude events. In the optical domain, rogue waves were first demonstrated experimentally in 2007 in the context of super-continuum generation in optical fibers and have since been observed in a wide variety of configurations such as semiconductor lasers. In QCLs, the extra power from these sudden bursts can be used in order to improve the efficiency of mid-infrared remote sensing or countermeasure systems. It can also be a helpful tool for neurophotonics clusters aiming to reproduce synaptic transmissions in an all-optical system. As a step toward a reliable control over these rare spikes, we carry out a statistical analysis of the interval between rogue events and show that precursors always occur before these events. The advantage of these precursors is to have a characteristic time longer than that found in other semiconductor lasers exhibiting the same non-linear phenomena. Birth of giant pulses like dragon-kings events are also discussed and analyzed. (10.1117/12.2545217)
  DOI : 10.1117/12.2545217
• Posterior consistency for partially observed Markov models
  
  • Douc Randal
  • Olsson Jimmy
  • Roueff François
  Stochastic Processes and their Applications, Elsevier, 2020, 130 (2), pp.733-759. We establish the posterior consistency for parametric, partially observed, fully dominated Markov models. The prior is assumed to assign positive probability to all neighborhoods of the true parameter, for a distance induced by the expected Kullback–Leibler divergence between the parametric family members’ Markov transition densities. This assumption is easily checked in general. In addition, we show that the posterior consistency is implied by the consistency of the maximum likelihood estimator. The result is extended to possibly improper priors and non-stationary observations. Finally, we check our assumptions on a linear Gaussian model and a well-known stochastic volatility model. (10.1016/j.spa.2019.03.012)
  DOI : 10.1016/j.spa.2019.03.012
• High-performance mode-locked lasers on silicon (invited)
  
  • Liu Songtao
  • Wu Xinru
  • Norman Justin
  • Jung Daehwan
  • Dumont Mario
  • Shang Chen
  • Wan Yating
  • Kennedy M.
  • Dong Bozhang
  • Auth Dominik
  • Breuer Stefan
  • Grillot Frédéric
  • Chow Weng
  • Gossard Arthur
  • Bowers John
  , 2020, pp.55. (10.1117/12.2552224)
  DOI : 10.1117/12.2552224
• Resilience by design & failures forecasting for a connected autonomous vehicle
  
  • Monteuuis Jean-Philippe
  , 2020. Autonomous vehicles with an automation level 5 will drive autonomously in any road scenarios such as highways, snowy roads, urban areas, or traffic jams. The integration of V2X communication, as a new source of perception for the vehicle could remove the limitations of local perception by communicating with an occluded pedestrian or by detecting in advance the presence of a vehicle under a heavy mist. However, this V2X communication may be a new source of attacks threatening the vehicle perception. Current countermeasures are not designed for all autonomous vehicles because these countermeasures require the driver assistance or work with a specific set of sensors. Therefore, the thesis aims to propose a generic failure resilient perception architecture for all types of connected and autonomous vehicles supporting different kinds of sensors. In this thesis, we propose a generic perception architecture named GPA with its failure resilient perception algorithm (FRPA). We propose a new threat analysis and risk assessment method named SARA that identifies and assess the risk of attacks targeting connected and automated vehicles with an automation level 5. To identify where and how these attacks occur, we propose an attacker and a security goal model for all automotive perception systems. We implemented two modules of our failures resilient perception algorithm (FRPA): a Machine Learning based Failure Classifier and a V2X-Sensor Correlation Module considering three kinds of source: camera, radar, and V2X. We highlighted several new attacks in the perception pipeline and raise the need for new security countermeasures such as the physical integrity of road infrastructures and trustworthy perception algorithms. Besides, our countermeasures based on machine learning and sensor correlation showed very accurate results to detect and classifies perception failures (over 90% accuracy score). Finally, the ideas developed in the thesis resulted in 10 filled patents and several publications.