Publications

2022

• InfoLM: A New Metric to Evaluate Summarization & Data2Text Generation
  
  • Colombo Pierre Jean A.
  • Clavel Chloé
  • Piantanida Pablo
  , 2022, 36 (10), pp.10554-10562. Assessing the quality of natural language generation (NLG) systems through human annotation is very expensive. Additionally, human annotation campaigns are time-consuming and include non-reusable human labour. In practice, researchers rely on automatic metrics as a proxy of quality. In the last decade, many string-based metrics (e.g., BLEU or ROUGE) have been introduced. However, such metrics usually rely on exact matches and thus, do not robustly handle synonyms. In this paper, we introduce InfoLM a family of untrained metrics that can be viewed as a string-based metric that addresses the aforementioned flaws thanks to a pre-trained masked language model. This family of metrics also makes use of information measures allowing the possibility to adapt InfoLM to different evaluation criteria. Using direct assessment, we demonstrate that InfoLM achieves statistically significant improvement and two figure correlation gains in many configurations compared to existing metrics on both summarization and data2text generation tasks. (10.1609/aaai.v36i10.21299)
  DOI : 10.1609/aaai.v36i10.21299
• Resource-Aware Edge-Based Stream Analytics
  
  • Petri Ioan
  • Chirila Ioan
  • Gomes Heitor Murilo
  • Bifet Albert
  • Rana Omer F.
  IEEE Internet Computing, Institute of Electrical and Electronics Engineers, 2022, 26 (4), pp.79--88. Understanding how machine learning (ML) algorithms can be used for stream processing on edge devices remains an important challenge. Such ML algorithms can be represented as operators and dynamically adapted based on the resources on which they are hosted. Deploying ML algorithms on edge resources often focuses on carrying out inference on the edge, while learning and model development takes place on a cloud data center. In this article, we describe TinyMOA, a modified version of the open-source massive online analytics library for stream processing, that can be deployed across both local and remote edge resources using the Parsl and Kafka systems. Using an experimental testbed, we demonstrate how ML stream-processing operators can be configured based on the resource on which they are hosted, and discuss subsequent implications for edge-based stream-processing systems. (10.1109/MIC.2022.3152478)
  DOI : 10.1109/MIC.2022.3152478
• Unsupervised anomaly detection : methods and applications
  
  • Putina Andrian
  , 2022. An anomaly (also known as outlier) is an instance that significantly deviates from the rest of the input data and being defined by Hawkins as 'an observation, which deviates so much from other observations as to arouse suspicions that it was generated by a different mechanism'. Anomaly detection (also known as outlier or novelty detection) is thus the machine learning and data mining field with the purpose of identifying those instances whose features appear to be inconsistent with the remainder of the dataset. In many applications, correctly distinguishing the set of anomalous data points (outliers) from the set of normal ones (inliers) proves to be very important. A first application is data cleaning, i.e., identifying noisy and fallacious measurement in a dataset before further applying learning algorithms. However, with the explosive growth of data volume collectable from various sources, e.g., card transactions, internet connections, temperature measurements, etc. the use of anomaly detection becomes a crucial stand-alone task for continuous monitoring of the systems. In this context, anomaly detection can be used to detect ongoing intrusion attacks, faulty sensor networks or cancerous masses.The thesis proposes first a batch tree-based approach for unsupervised anomaly detection, called 'Random Histogram Forest (RHF)'. The algorithm solves the curse of dimensionality problem using the fourth central moment (aka kurtosis) in the model construction while boasting linear running time. A stream based anomaly detection engine, called 'ODS', that leverages DenStream, an unsupervised clustering technique is presented subsequently and finally Automated Anomaly Detection engine which alleviates the human effort required when dealing with several algorithm and hyper-parameters is presented as last contribution.
• Graph-based contributions to machine-learning
  
  • Lutz Quentin
  , 2022. A graph is a mathematical object that makes it possible to represent relationships (called edges) between entities (called nodes). Graphs have long been a focal point in a number of problems ranging from work by Euler to PageRank and shortest-path problems. In more recent times, graphs have been used for machine learning.With the advent of social networks and the world-wide web, more and more datasets can be represented using graphs. Those graphs are ever bigger, sometimes with billions of edges and billions of nodes. Designing efficient algorithms for analyzing those datasets has thus proven necessary. This thesis reviews the state of the art and introduces new algorithms for the clustering and the embedding of the nodes of massive graphs. Furthermore, in order to facilitate the handling of large graphs and to apply the techniques under study, we introduce Scikit-network, a free and open-source Python library which was developed during the thesis. Many tasks, such as the classification or the ranking of the nodes using centrality measures, can be carried out thanks to Scikit-network.We also tackle the problem of labeling data. Supervised machine learning techniques require labeled data to be trained. The quality of this labeled data has a heavy influence on the quality of the predictions of those techniques once trained. However, building this data cannot be achieved through the sole use of machines and requires human intervention. We study the data labeling problem in a graph-based setting, and we aim at describing the solutions that require as little human intervention as possible. We characterize those solutions and illustrate how they can be applied in real use-cases.
• Conception et intégration d'un convertisseur analogique-paramètres flexible pour les capteurs intelligents
  
  • Back Antoine
  , 2022. Avec le fort développement de l'Internet des Objets (IoT), il devient nécessaire de converger vers de nouveaux capteurs dit intelligents. Ces capteurs doivent permettre d'analyser l'environnement extérieur, comprendre le contexte dans lequel ils sont utilisés et être conscient des besoins utilisateurs. Ils doivent cependant rester petits, fiables, bon marché et avoir une autonomie de plusieurs années. La conversion analogique-paramètre (Analog-to-Feature, A2F) est une nouvelle méthode d'acquisition pensée pour les appareils IoT, et semble être une solution adaptée pour de tels capteurs. Cette conversion consiste à extraire des paramètres directement sur le signal analogique. Une sélection pertinente des paramètres permet d'extraire uniquement l'information nécessaire à une tache particulière. Le convertisseur proposé est basé sur la technique de l'échantillonnage non-uniforme en ondelettes (NUWS). L'architecture mélange le signal analogique avec des ondelettes paramétrables avant d'intégrer et convertir le signal en données numériques. L'objectif de la thèse est de proposer une méthode pour concevoir un convertisseur A2F générique basé sur le NUWS. Il est ainsi nécessaire de définir les caractéristiques des ondelettes afin d'acquérir une large gamme de signaux basse fréquence (ECG, EMG, EEG, parole…). Cette étape nécessite l'utilisation d'algorithmes de sélection de paramètres et d'algorithmes d'apprentissage automatique pour sélectionner le meilleur ensemble d'ondelettes pour une application donnée et qui doit permettre de définir les spécifications du convertisseur. L'étape de sélection des paramètres doit tenir compte des contraintes de mise en œuvre pour optimiser au mieux la consommation d'énergie. Un algorithme de sélection de paramètres est proposé pour choisir des ondelettes pour une application donnée, afin de maximiser la précision de classification tout en diminuant la consommation d'énergie, grâce à un modèle de consommation réalisé dans une technologie CMOS 0.18μm.
• SysML Models Verification Relying on Dependency Graphs
  
  • Apvrille Ludovic
  • de Saqui-Sannes Pierre
  • Hotescu Oana
  • Calvino Alessandro Tempia
  , 2022, 1, pp.174-181. Formal verification of SysML models contributes to early detection of design errors early in the life cycle of systems. Incremental modeling of systems leads to the repeated verification of parts of systems models that were already verified in previous versions of the SysML model. This paper proposes to optimize the verification process by generating dependency graphs from SysML models. This revisited verification technique is now supported by TTool. It is illustrated with an Avionics Full DupleX network. (10.5220/0010792900003119)
  DOI : 10.5220/0010792900003119
• AutoML: state of the art with a focus on anomaly detection, challenges, and research directions
  
  • Bahri Maroua
  • Salutari Flavia
  • Putina Andrian
  • Sozio Mauro
  International Journal of Data Science and Analytics, Springer Verlag, 2022. The last decade has witnessed the explosion of machine learning research studies with the inception of several algorithms proposed and successfully adopted in different application domains. However, the performance of multiple machine learning algorithms is very sensitive to multiple ingredients (e.g., hyper-parameters tuning and data cleaning) where a significant human effort is required to achieve good results. Thus, building well-performing machine learning algorithms requires domain knowledge and highly specialized data scientists. Automated machine learning (autoML) aims to make easier and more accessible the use of machine learning algorithms for researchers with varying levels of expertise. Besides, research effort to date has mainly been devoted to autoML for supervised learning, and only a few research proposals have been provided for the unsupervised learning. In this paper, we present an overview of the autoML field with a particular emphasis on the automated methods and strategies that have been proposed for unsupervised anomaly detection. (10.1007/s41060-022-00309-0)
  DOI : 10.1007/s41060-022-00309-0
• High bandwidth detection of mechanical stress in optical fibre using coherent detection of Rayleigh scattering
  
  • Guerrier Sterenn
  , 2022. Telecommunication fibres are being deployed all over the world, connecting distant people, institutions, companies with an outstanding quality of service in terms of data rate and latency. Their strategic value in terms of global economy and daily life is now undeniable. Monitoring such an infrastructure has become mandatory, and that far beyond the standard case of breaks localization. From a broader standpoint, optical fibres are an alternative to electro-dynamic point sensors, with a strong asset: the capability to detect and localize multiple independent phenomena all along a fibre. Thus, the millions of kilometers of currently deployed optical fibre around the world constitute a huge potential base of sensors. Distributed vibration sensors have a huge potential regarding sensing of dynamic events, detecting of multiple acoustic signatures up to speech signals. In this thesis, we show how distributed optical fibre sensors can be designed on top of telecommunication fibres, namely standard single mode fibres, and we explore their potential in terms of reach, detection threshold, and sensing bandwidth. We present the interrogator systems for distributed fibre sensing and build a dual-polarization numerical model of such an interrogation system and fibre sensor. Secondly, we tackle the coherent fading issue by means of frequency diversity in the digital domain, i.e. directly applicable at the modulation of the interrogation sequences, before entering the optical domain. We developed MIMO-OFDM which retrieves independent channel estimations from a single fibre segment; the estimations are further combined, and the obtained estimations are assessed with regards to the reliability metric. Throughout this thesis, many experimental measurements were conducted, assessing the capabilities of the Coherent-MIMO interrogator on single-mode-fibre sensors in terms of reach, bandwidth, and detection threshold. We also demonstrate the co-propagation of a sensing signal along with high data rate channels, without any impact on the transmitted data, paving the way to the enhanced monitoring and telemetry in deployed telecommunication networks.
• Enabling Markovian Representations under Imperfect Information
  
  • Belardinelli Francesco
  • G. León Borja
  • Malvone Vadim
  , 2022, 2, pp.450-457. Markovian systems are widely used in reinforcement learning (RL), when the successful completion of a task depends exclusively on the last interaction between an autonomous agent and its environment. Unfortunately, real-world instructions are typically complex and often better described as non-Markovian. In this paper we present an extension method that allows solving partially-observable non-Markovian reward decision processes (PONMRDPs) by solving equivalent Markovian models. This potentially facilitates Markovian-based state-of-the-art techniques, including RL, to find optimal behaviours for problems best described as PONMRDP. We provide formal optimality guarantees of our extension methods together with a counterexample illustrating that naive extensions from existing techniques in fully-observable environments cannot provide such guarantees. (10.5220/0010882200003116)
  DOI : 10.5220/0010882200003116
• Wireless phone use in childhood and adolescence and neuroepithelial brain tumours: Results from the international MOBI-Kids study
  
  • Castaño-Vinyals G.
  • Sadetzki S.
  • Vermeulen R.
  • Momoli F.
  • Kundi M.
  • Merletti F.
  • Maslanyj M.
  • Calderon C.
  • Wiart Joe
  • Lee A.-K.
  • Taki M.
  • Sim M.
  • Armstrong B.
  • Benke G.
  • Schattner R.
  • Hutter H.-P.
  • Krewski D.
  • Mohipp C.
  • Ritvo P.
  • Spinelli J.
  • Lacour B.
  • Remen T.
  • Radon K.
  • Weinmann T.
  • Petridou E.Th.
  • Moschovi M.
  • Pourtsidis A.
  • Oikonomou K.
  • Kanavidis P.
  • Bouka E.
  • Dikshit R.
  • Nagrani R.
  • Chetrit A.
  • Bruchim R.
  • Maule M.
  • Migliore E.
  • Filippini G.
  • Miligi L.
  • Mattioli S.
  • Kojimahara N.
  • Yamaguchi N.
  • Ha M.
  • Choi K.
  • Kromhout H.
  • Goedhart G.
  • T Mannetje A.
  • Eng A.
  • Langer C.E.
  • Alguacil J.
  • Aragonés N.
  • Morales-Suárez-Varela M.
  • Badia F.
  • Albert A.
  • Carretero G.
  • Cardis E.
  Environment International, Elsevier, 2022, 160, pp.107069. (10.1016/j.envint.2021.107069)
  DOI : 10.1016/j.envint.2021.107069
• Saturday knights
  
  • Boyer Ivan
  • Zayana Karim
  • Rabiet Victor
  CultureMath, ENS, 2022.
• The consensus number of a cryptocurrency
  
  • Guerraoui Rachid
  • Kuznetsov Petr
  • Monti Matteo
  • Pavlovic Matej
  • Seredinschi Dragos-Adrian
  Distributed Computing, Springer Verlag, 2022, 35 (1), pp.1-15. Many blockchain-based algorithms, such as Bitcoin, implement a decentralized asset transfer system, often referred to as a cryptocurrency. As stated in the original paper by Nakamoto, at the heart of these systems lies the problem of preventing double-spending; this is usually solved by achieving consensus on the order of transfers among the participants. In this paper, we treat the asset transfer problem as a concurrent object and determine its consensus number, showing that consensus is, in fact, not necessary to prevent double-spending. We first consider the problem as defined by Nakamoto, where only a single process—the account owner—can withdraw from each account. Safety and liveness need to be ensured for correct account owners, whereas misbehaving account owners might be unable to perform transfers. We show that the consensus number of an asset transfer object is 1. We then consider a more general k -shared asset transfer object where up to k processes can atomically withdraw from the same account, and show that this object has consensus number k . We establish our results in the context of shared memory with benign faults, allowing us to properly understand the level of difficulty of the asset transfer problem. We also translate these results in the message passing setting with Byzantine players, a model that is more relevant in practice. In this model, we describe an asynchronous Byzantine fault-tolerant asset transfer implementation that is both simpler and more efficient than state-of-the-art consensus-based solutions. Our results are applicable to both the permissioned (private) and permissionless (public) setting, as normally their differentiation is hidden by the abstractions on top of which our algorithms are based. (10.1007/s00446-021-00399-2)
  DOI : 10.1007/s00446-021-00399-2
• Assessment and Mitigation of Power Side-Channel-Based Cross-PUF Attacks on Arbiter-PUFs and Their Derivatives
  
  • Kroeger Trevor
  • Cheng Wei
  • Guilley Sylvain
  • Danger Jean-Luc
  • Karimi Naghmeh
  IEEE Transactions on Very Large Scale Integration (VLSI) Systems, IEEE, 2022, 30 (2), pp.187-200. Unintentional uncontrollable variations in the manufacturing process of integrated circuits are used to realize silicon primitives known as physical unclonable functions (PUFs). These primitives are used to create unique signatures for security purposes. Investigating the vulnerabilities of PUFs is of utmost importance to uphold their usefulness in secure applications. One such investigation includes exploring the susceptibility of PUFs to modeling attacks that aim at extracting the PUFs’ behavior. To date, these attacks have mainly focused on a single PUF instance where the targeted PUF is attacked using the model built based on the very same PUF’s challenge–response pairs or power side channel. In this article, we move one step forward and introduce Cross-PUF attacks where a model is created using the power consumption of one PUF instance to attack another PUF created from the same GDSII file. Through SPICE simulations, we show that these attacks are highly effective in modeling PUF behaviors even in the presence of noise and mismatches in temperature and aging of the PUF used for modeling versus the targeted PUF. To mitigate the Cross-PUF attacks, we then propose a lightweight countermeasure based on dual-rail and random initialization logic approaches called DRILL. We show that DRILL is highly effective in thwarting Cross-PUF attacks (10.1109/TVLSI.2021.3129141)
  DOI : 10.1109/TVLSI.2021.3129141
• Joint Content-prefetching, Transmission Scheduling,and Rate Adaptation in Vehicular Networks
  
  • Berri Sara
  • Zhang Jun
  • Bensaou Brahim
  • Labiod Houda
  IEEE Transactions on Vehicular Technology, Institute of Electrical and Electronics Engineers, 2022.
• Esthétique de la photographie numérique, ISTE Editions, ISBN: 978-1-78405-815-9
  
  • Maître Henri
  , 2022.
• On those multiplicative subgroups of $${\mathbb F}_{2^n}^*$$ which are Sidon sets and/or sum-free sets
  
  • Carlet Claude
  • Mesnager Sihem
  Journal of Algebraic Combinatorics, Springer Verlag, 2022, 55 (1), pp.43-59. (10.1007/s10801-020-00988-7)
  DOI : 10.1007/s10801-020-00988-7
• 10 Gbit s −1 Free Space Data Transmission at 9 µm Wavelength With Unipolar Quantum Optoelectronics
  
  • Dely Hamza
  • Bonazzi Thomas
  • Spitz Olivier
  • Rodriguez Etienne
  • Gacemi Djamal
  • Todorov Yanko
  • Pantzas Konstantinos
  • Beaudoin Grégoire
  • Sagnes Isabelle
  • Li Lianhe
  • Davies Alexander Giles
  • Linfield Edmund
  • Grillot Frédéric
  • Vasanelli Angela
  • Sirtori Carlo
  Laser and Photonics Reviews, Wiley-VCH Verlag, 2022, 16 (2), pp.2100414. Free space optics data transmission with bitrate in excess of 10 Gbit s−1 is demonstrated at 9 µm wavelength by using a unipolar quantum optoelectronic system at room temperature, composed of a quantum cascade laser, a modulator, and a quantum cascade detector. The large frequency bandwidth of the system is set by the detector and the modulator that are both high frequency devices, while the laser emits in continuous wave. The amplitude modulator relies on the Stark shift of an absorbing optical transition in and out of the laser frequency. This device is designed to avoid charge displacement, and therefore it is characterized by an intrinsically large bandwidth and very low electrical power consumption. This demonstration of high-bitrate data transmission sets unipolar quantum devices as the most performing platform for the development of optoelectronic systems operating at very high frequency in the mid-infrared for several applications, such as digital communications and high-resolution spectroscopy. (10.1002/lpor.202100414)
  DOI : 10.1002/lpor.202100414
• LOss-Based SensiTivity rEgulaRization: Towards deep sparse neural networks
  
  • Tartaglione Enzo
  • Bragagnolo Andrea
  • Fiandrotti Attilio
  • Grangetto Marco
  Neural Networks, Elsevier, 2022, 146, pp.230-237. (10.1016/j.neunet.2021.11.029)
  DOI : 10.1016/j.neunet.2021.11.029
• Hidden regular variation for point processes and the single/multiple large point heuristic
  
  • Dombry Clément
  • Wintenberger Olivier
  • Tillier Charles
  The Annals of Applied Probability, Institute of Mathematical Statistics (IMS), 2022, 32 (1). We consider regular variation for marked point processes with independent heavy-tailed marks and prove a single large point heuristic: the limit measure is concentrated on the cone of point measures with one single point. We then investigate successive hidden regular variation removing the cone of point measures with at most k points, k ≥ 1, and prove a multiple large point phenomenon: the limit measure is concentrated on the cone of point measures with k + 1 points. We show how these results imply hidden regular variation in Sko-rokhod space of the associated risk process, in connection with the single/multiple large point heuristic from Rhee et al. (2019). Finally, we provide an application to risk theory in a reinsurance model where the k largest claims are covered and we study the asymptotic behavior of the residual risk. (10.1214/21-AAP1675)
  DOI : 10.1214/21-AAP1675
• Novel Centralized Pseudonym Changing Scheme for Location Privacy in V2X Communication
  
  • Didouh Ahmed
  • El Hillali Yassin
  • Rivenq Atika
  • Labiod Houda
  Energies, MDPI, 2022, 15 (3), pp.692. Vehicular ad hoc networks allow vehicles to share their information for the safety and efficiency of traffic purposes. However, information sharing can threaten the driver’s privacy as it includes spatiotemporal information, and the messages are unencrypted and broadcasted periodically. Therefore, they cannot estimate their privacy level because it also depends on their surroundings. This article proposes a centralized adaptive pseudonym change scheme that permits the certificate’s authority to adjust the pseudonyms assignment for each requesting vehicle. This scheme adapts dynamically depending on the density of the traffic environment and the user’s privacy level, and it aims to solve the trade-off problem between wasting pseudonyms and Sybil attack. We employ a Knapsack problem-based algorithm for target tracking and an entropy-based method to measure each vehicle’s privacy. In order to demonstrate the applicability of our framework, we use real-life data captured during the interoperability tests of the European project InterCor. According to the experimental results, the proposed scheme could easily estimate the level of confidentiality and, therefore, may best respond to the adaptation of the pseudonyms. (10.3390/en15030692)
  DOI : 10.3390/en15030692
• A SIC Based BS Coordination Scheme for Full Duplex Cellular Networks
  
  • Coupechoux Marceau
  • Arraño-Scharager Hernán-Felipe
  • Kelif Jean-Marc
  IEEE Transactions on Communications, Institute of Electrical and Electronics Engineers, 2022, 70 (2), pp.1043-1057. Full Duplex (FD) in cellular networks is expected to increase the cell spectral efficiency. However, while the downlink (DL) spectral efficiency (SE) increases with FD, the uplink (UL) SE decreases because of the Base Station to Base Station (BS) interference. In this paper, assuming a three-node model, we propose a method based on Successive Interference Cancellation (SIC) to reduce the BS-to-BS interference present in FD cellular networks. The approach consists in coordinating BSs to enable the decoding and the suppression of undesired signals that impair uplink transmissions. We analyze both distributed and Centralized Radio Access Networks (CRAN) architectures. Stochastic geometry is used to derive the coverage probability and mean data rate of the proposed scheme. In the distributed scenario, the FD UL average data rate is increased by 25% with our solution compared to a classical FD network, while our FD scheme still outperforms Half-Duplex (HD) on the DL. In the centralized scenario, our solution outperforms HD by 10% and classical FD by 78% on the UL, while preserving classical FD gains on the DL. (10.1109/TCOMM.2021.3122919)
  DOI : 10.1109/TCOMM.2021.3122919
• Online Learning for Adaptive Video Streaming in Mobile Networks
  
  • Karagkioules Theodoros
  • Paschos Georgios
  • Liakopoulos Nikolaos
  • Fiandrotti Attilio
  • Tsilimantos Dimitrios
  • Cagnazzo Marco
  ACM Transactions on Multimedia Computing, Communications and Applications, Association for Computing Machinery, 2022, 18 (1), pp.1-22. In this paper, we propose a novel algorithm for video bitrate adaptation in HTTP Adaptive Streaming (HAS), based on online learning. The proposed algorithm, named Learn2Adapt (L2A) , is shown to provide a robust bitrate adaptation strategy which, unlike most of the state-of-the-art techniques, does not require parameter tuning, channel model assumptions, or application-specific adjustments. These properties make it very suitable for mobile users, who typically experience fast variations in channel characteristics. Experimental results, over real 4G traffic traces, show that L2A improves on the overall Quality of Experience (QoE) and in particular the average streaming bitrate, a result obtained independently of the channel and application scenarios. (10.1145/3460819)
  DOI : 10.1145/3460819
• Near‐Field Exposure in FM Frequencies: New Methodology and Estimation Formulas
  
  • Fetouri Bader Mustafa
  • Azzi Soumaya
  • Ouberehil Allal
  • Briend Philippe
  • de Doncker Philippe
  • Wiart Joe
  Bioelectromagnetics, Wiley, 2022. (10.1002/bem.22391)
  DOI : 10.1002/bem.22391
• The life and work of Kolmogorov
  
  • Rioul Olivier
  , 2022, pp.https://culturemath.ens.fr/thematiques/biographie/life-and-work-kolmogorov. Andrei Kolmogorov is one of the greatest mathematicians of the 20th century. He revolutionized every subject he approached from a surprisingly original perspective with astonishing insight and imagination...
• The Kingsguard OS-level mitigation against cache side-channel attacks using runtime detection
  
  • Mushtaq Maria
  • Yousaf Muhammad Muneeb
  • Bhatti Muhammad Khurram
  • Lapotre Vianney
  • Guy Gogniat
  Annals of Telecommunications - annales des télécommunications, Springer, 2022, 77, pp.731–747. Most of the mitigation techniques against access-driven cache side-channel attacks (CSCAs) are not very effective. This is mainly because most mitigation techniques usually protect against any given specific vulnerability of the system and do not take a system-wide approach. Moreover, they either completely remove or greatly reduce the performance benefits. Therefore, to find a security vs performance trade-off, we argue in favor of need-based protection in this paper, which will allow the operating system to apply mitigation only after successful detection of CSCAs. Thus, detection can serve as a first line of defense against such attacks. In this work, we propose a novel OS-level runtime detection-based mitigation mechanism, called the Kingsguard, against CSCAs in general-purpose operating systems. The proposed mechanism enhances the security and privacy capabilities of Linux as a proof of concept, and it can be widely used in commodity systems without any hardware modifications. We provide experimental validation by mitigating three state-of-the-art CSCAs on two different cryptosystems running under Linux. We have also provided results by analyzing the effect of the combination of multiple attacks running concurrently under variable system noise. Our results show that the Kingsguard can detect and mitigate known CSCAs with an accuracy of more than 99% and 95%, respectively. (10.1007/s12243-021-00906-3)
  DOI : 10.1007/s12243-021-00906-3