Direction de la Recherche Technologique

This paper presents a new classification framework for brain-computer interface (BCI) based on motor imagery. This framework involves the concept of Riemannian geometry in the manifold of covariance matrices. The main idea is to use spatial covariance matrices as EEG signal descriptors and to rely on Riemannian geometry to directly classify these matrices using the topology of the manifold of symmetric and positive definite (SPD) matrices. This framework allows to extract the spatial information contained in EEG signals without using spatial filtering. Two methods are proposed and compared with a reference method [multiclass Common Spatial Pattern (CSP) and Linear Discriminant Analysis (LDA)] on the multiclass dataset IIa from the BCI Competition IV. The first method, named minimum distance to Riemannian mean (MDRM), is an implementation of the minimum distance to mean (MDM) classification algorithm using Riemannian distance and Riemannian mean. This simple method shows comparable results with the reference method. The second method, named tangent space LDA (TSLDA), maps the covariance matrices onto the Riemannian tangent space where matrices can be vectorized and treated as Euclidean objects. Then, a variable selection procedure is applied in order to decrease dimensionality and a classification by LDA is performed. This latter method outperforms the reference method increasing the mean classification accuracy from 65.1% to 70.2%.

Abstract Resistive switching (RS) is an interesting property shown by some materials systems that, especially during the last decade, has gained a lot of interest for the fabrication of electronic devices, with electronic nonvolatile memories being those that have received the most attention. The presence and quality of the RS phenomenon in a materials system can be studied using different prototype cells, performing different experiments, displaying different figures of merit, and developing different computational analyses. Therefore, the real usefulness and impact of the findings presented in each study for the RS technology will be also different. This manuscript describes the most recommendable methodologies for the fabrication, characterization, and simulation of RS devices, as well as the proper methods to display the data obtained. The idea is to help the scientific community to evaluate the real usefulness and impact of an RS study for the development of RS technology.

Dynamic spectrum policies combined with software defined radio are powerful means to improve the overall spectral efficiency allowing the development of new wireless services and technologies. Medium Access Control (MAC) protocols exploit sensing stimuli to build up a spectrum opportunity map (cognitive sensing). Available resources are scheduled (dynamic spectrum allocation), improving coexistence between users that belong to heterogeneous systems (dynamic spectrum sharing). Furthermore, MAC protocols may allow cognitive users to vacate selected channels when their quality becomes unacceptable (dynamic spectrum mobility). The contribution of this survey is threefold. First, we show the fundamental role of the MAC layer and identify its functionalities in a cognitive radio (CR) network. Second, a classification of cognitive MAC protocols is proposed. Third, advantages, drawbacks, and further design challenges of cognitive MAC protocols are discussed.

Energy efficiency is the key requirement to maximize sensor node lifetime. Sensor nodes are typically powered by a battery source that has finite lifetime. Most Internet of Thing (IoT) applications require sensor nodes to operate reliably for an extended period of time. To design an autonomous sensor node, it is important to model its energy consumption for different tasks. Each task consumes a power consumption amount for a period of time. To optimize the consumed energy of the sensor node and have long communication range, Low Power Wide Area Network technology is considered. This paper describes an energy consumption model based on LoRa and LoRaWAN, which allows estimating the consumed power of each sensor node element. The definition of the different node units is first introduced. Then, a full energy model for communicating sensors is proposed. This model can be used to compare different LoRaWAN modes to find the best sensor node design to achieve its energy autonomy.

Practical and theoretical considerations that apply when aiming to formulate by ultrasonication very small nanoemulsions (particle diameter up to 150 nm) with very high stability are presented and discussed. The droplet size evolution during sonication can be described by a monoexponential function of the sonication time, the characteristic time scale depending essentially on the applied power. A unique master curve is obtained when plotting the mean diameter size evolution as a function of sonication energy. We then show that Ostwald ripening remains the main destabilization mechanism whereas coalescence can be easily prevented due to the nanometric size of droplets. The incorporation of "trapped species" within the droplet interior is able to counteract Ostwald ripening, and this concept can be extended to the membrane compartment. We finally clarify that nanoemulsions are not thermodynamically stable systems, even in the case where their composition lies very close to the demixing line of a thermodynamically stable microemulsion domain. However, as exemplified in the present work, nanoemulsion systems can present very long-term kinetic stability.

During the postnatal development, astrocytic cells in the neocortex progressively lose their neural stem cell (NSC) potential, whereas this peculiar attribute is preserved in the adult subventricular zone (SVZ). To understand this fundamental difference, many reports suggest that adult subventricular GFAP-expressing cells might be maintained in immature developmental stage. Here, we show that S100B, a marker of glial cells, is absent from GFAP-expressing cells of the SVZ and that its onset of expression characterizes a terminal maturation stage of cortical astrocytic cells. Nevertheless, when cultured in vitro, SVZ astrocytic cells developed as S100B expressing cells, as do cortical astrocytic cells, suggesting that SVZ microenvironment represses S100B expression. Using transgenic s100b-EGFP cells, we then demonstrated that S100B expression coincides with the loss of neurosphere forming abilities of GFAP expressing cells. By doing grafting experiments with cells derived from beta-actin-GFP mice, we next found that S100B expression in astrocytic cells is repressed in the SVZ, but not in the striatal parenchyma. Furthermore, we showed that treatment with epidermal growth factor represses S100B expression in GFAP-expressing cells in vitro as well as in vivo. Altogether, our results indicate that the S100B expression defines a late developmental stage after which GFAP-expressing cells lose their NSC potential and suggest that S100B expression is repressed by adult SVZ microenvironment.

We propose a modification of Wells et al. technique for bias field estimation and segmentation of magnetic resonance (MR) images. We show that replacing the class other, which includes all tissue not modeled explicitly by Gaussians with small variance, by a uniform probability density, and amending the expectation-maximization (EM) algorithm appropriately, gives significantly better results. We next consider the estimation and filtering of high-frequency information in MR images, comprising noise, intertissue boundaries, and within tissue microstructures. We conclude that post-filtering is preferable to the prefiltering that has been proposed previously. We observe that the performance of any segmentation algorithm, in particular that of Wells et al. (and our refinements of it) is affected substantially by the number and selection of the tissue classes that are modeled explicitly, the corresponding defining parameters and, critically, the spatial distribution of tissues in the image. We present an initial exploration to choose automatically the number of classes and the associated parameters that give the best output. This requires us to define what is meant by "best output" and for this we propose the application of minimum entropy. The methods developed have been implemented and are illustrated throughout on simulated and real data (brain and breast MR).

This paper investigates terahertz detectors fabricated in a low-cost 130 nm silicon CMOS technology. We show that the detectors consisting of a nMOS field effect transistor as rectifying element and an integrated bow-tie coupling antenna achieve a record responsivity above 5 kV/W and a noise equivalent power below 10 pW/Hz(0.5) in the important atmospheric window around 300 GHz and at room temperature. We demonstrate furthermore that the same detectors are efficient for imaging in a very wide frequency range from ~0.27 THz up to 1.05 THz. These results pave the way towards high sensitivity focal plane arrays in silicon for terahertz imaging.

Part 1 How to use a computer algebra system introduction features of computer algebra systems syntax of the associated languages areas covered by existing systems computer algebra by example MACSYMA's possibilities in algebra general possibilities the division of the circle into 17 equal parts availability of MACSYMA other systems AXIOM. Part 2 The problem of data representation representations of data representation representations of integers representations of fractions, representations of polynomials polynomials in several variables representations of rational functions representations of algebraic functions representations of transcendentals representations of matrices, representations of series. Part 3 Polynomial simplification simplification of polynomial equations simplification of real polynomial systems. Part 4 Advanced algorithms modular methods p-adic methods. Part 5 Formal integration and differential equations introduction integration of rational functions the integration of more complicated functions integration of logarithmic functions integration of exponential functions, algebraic solutions of o.d.e.s. asymptotic solutions of o.d.e.s. appendix algebraic background annex REDUCE - a computer algebra system introduction syntax of REDUCE declarations commands built-in facilities manipulation of expressions substitution matrix algebra IRENA.

A 2D map is created for solid-state materials based on a quantum-mechanical description of electron sharing and electron transfer. This map intuitively identifies the fundamental nature of ionic, metallic, and covalent bonding in a range of elements and binary compounds; furthermore, it highlights a distinct region for a mechanism recently termed "metavalent" bonding. Then, it is shown how this materials map can be extended in the third dimension by including physical properties of application interest. Finally, it is shown how the map coordinates yield new insight into the nature of the Peierls distortion in phase-change materials and thermoelectrics. These findings and conceptual approaches provide a novel avenue to tailor material properties.

GaN-based micro light-emitting diode (μLED) arrays are very promising devices for display applications. In these arrays, each μLED works as a single pixel of a whole image. The electro-optical performance of these μLEDs is an important subject to study. Here, we investigate the influence of LED size on the radiative and non-radiative recombination. The standard ABC model has been widely used to describe the efficiency of GaN based LEDs. Using this model, we extract A, B, and C coefficients for various LED sizes, showing how the competition between radiative and non-radiative recombination processes varies with the LED geometry. Time-resolved photoluminescence allows us to determine coefficient B, related to radiative recombination. Through current-voltage-luminance characterizations, we determine parameters A and C related to Shockley-Read-Hall and Auger recombination. We find that coefficient A is strongly dependent on LED size, indicating a drastic effect of sidewall defects on the performance of LEDs. On the other hand, coefficient C is independent of LED size. This latter result demonstrates that efficiency droop does not depend on LED size.

22FDX™ is the industry's first FDSOI technology architected to meet the requirements of emerging mobile, Internet-of-Things (IoT), and RF applications. This platform achieves the power and performance efficiency of a 16/14nm FinFET technology in a cost effective, planar device architecture that can be implemented with ~30% fewer masks. Performance comes from a second generation FDSOI transistor, which produces nFET (pFET) drive currents of 910μ/μm (856μ/μm) at 0.8 V and 100nA/μm Ioff. For ultra-low power applications, it offers low-voltage operation down to 0.4V V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">min</sub> for 8T logic libraries, as well as 0.62V and 0.52V V <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">min</sub> for high-density and high-current bitcells, ultra-low leakage devices approaching 1pA/μm I <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">off</sub> , and body-biasing to actively trade-off power and performance. Superior RF/Analog characteristics to FinFET are achieved including high f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</sub> /f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MAx</sub> of 375GHz/290GHz and 260GHz/250GHz for nFET and pFET, respectively. The high f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MAx</sub> extends the capabilities to 5G and milli-meter wave (>24GHz) RF applications.

The solubilizing power of various nonionic and zwitterionic detergents as membrane protein solubilizers for two-dimensional electrophoresis was investigated. Human red blood cell ghosts and Arabidopsis thaliana leaf membrane proteins were used as model systems. Efficient detergents could be found in each class, i.e. with oligooxyethylene, sugar or sulfobetaine polar heads. Among the commercially available nonionic detergents, dodecyl maltoside and decaethylene glycol mono hexadecyl ether proved most efficient. They complement the more classical sulfobetaine detergents to widen the scope of useful detergents for the solubilization of membrane proteins in proteomics.

Protein kinase CK2 is a ubiquitous protein kinase implicated in proliferation and cell survival. Its regulatory beta subunit, CK2beta, which is encoded by a single gene in mammals, has been suspected of regulating other protein kinases. In this work, we show that knockout of the CK2beta gene in mice leads to postimplantation lethality. Mutant embryos were reduced in size at embryonic day 6.5 (E6.5). They did not exhibit signs of apoptosis but did show reduced cell proliferation. Mutant embryos were resorbed at E7.5. In vitro, CK2beta(-/-) morula development stopped after the blastocyst stage. Attempts to generate homozygous embryonic stem (ES) cells failed. By using a conditional knockout approach, we show that lack of CK2beta is deleterious for mouse ES cells and primary embryonic fibroblasts. This is in contrast to what occurs with yeast cells, which can survive without functional CK2beta. Thus, our study demonstrates that in mammals, CK2beta is essential for viability at the cellular level, possibly because it acquired new functions during evolution.

The Industrial Internet of Things (IIoT) is an emerging technology that can promote the development of industrial intelligence, improve production efficiency, and reduce manufacturing costs. However, anomalies of IIoT devices might expose sensitive data about users of high authenticity and validity, resulting in security and privacy threats to the IIoT applications. That suggests the significance of anomaly detection executed by proper authorities. To address these problems, in this paper, we propose a reliable anomaly detection strategy for IIoT using federated learning. Specifically, we apply the federated learning technique to build a universal anomaly detection model with each local model trained by the deep reinforcement learning (DRL) algorithm. Since local data sets are not required during the federated learning, the chance of privacy leakage is reduced. In addition, by introducing privacy leakage degree and action relation to anomaly detection design, we can greatly improve the detection accuracy. The validation experiments indicate that the proposed strategy achieves high throughput, low latency, and high anomaly detection accuracy for privacy preservation in various IIoT scenarios.

Ge-based photodetectors operating in the low loss windows (1.3–1.6 μm) of silica fibers are highly desirable for the development of optical interconnections on silicon-on-insulator substrates. We have therefore investigated the structural and optical properties of Ge thick films grown directly onto Si(001) substrates using a production-compatible reduced pressure chemical vapor deposition system. We have first of all evidenced a Ge growth regime which is akin to a supply-limited one in the 400–750 °C temperature range (Ea=6.9 kcal mol−1). The thick Ge layers grown using a low-temperature/high-temperature approach are in a definite tensile-strain configuration, with a threading dislocation density for as-grown layers of the order of 9×108 cm−2 (annealed: &amp;lt;2×108 cm−2). The surface of those Ge thick layers is rather smooth, especially when considering the large lattice mismatch between Ge and Si. The root-mean-square roughness is indeed of the order of 0.6 nm (2 nm) only for as-grown (annealed) layers. A chemical mechanical polishing step followed by some Ge re-epitaxy can help in bringing the surface roughness of annealed layers down, however (0.5 nm). The Ge layers produced are of high optical quality. An absorption coefficient alpha equal to 4300 cm−1 (3400 cm−1) has indeed been found at room temperature and for a 1.55-μm wavelength for as-grown (annealed) layers. A 20-meV band-gap shrinkage with respect to bulk Ge (0.78 eV⇔0.80 eV) is observed as well in those tensile-strained Ge epilayers.

Abstract This roadmap presents the transformational research ideas proposed by “BATTERY 2030+,” the European large‐scale research initiative for future battery chemistries. A “chemistry‐neutral” roadmap to advance battery research, particularly at low technology readiness levels, is outlined, with a time horizon of more than ten years. The roadmap is centered around six themes: 1) accelerated materials discovery platform, 2) battery interface genome, with the integration of smart functionalities such as 3) sensing and 4) self‐healing processes. Beyond chemistry related aspects also include crosscutting research regarding 5) manufacturability and 6) recyclability. This roadmap should be seen as an enabling complement to the global battery roadmaps which focus on expected ultrahigh battery performance, especially for the future of transport. Batteries are used in many applications and are considered to be one technology necessary to reach the climate goals. Currently the market is dominated by lithium‐ion batteries, which perform well, but despite new generations coming in the near future, they will soon approach their performance limits. Without major breakthroughs, battery performance and production requirements will not be sufficient to enable the building of a climate‐neutral society. Through this “chemistry neutral” approach a generic toolbox transforming the way batteries are developed, designed and manufactured, will be created.

Photonic integrated circuits offer the potential of realizing low-cost, compact optical functions. Silicon-on-insulator (SOI) is a promising material platform for this photonic integration, as one can rely on the massive electronics processing infrastructure to process the optical components. However, the integration of a Si laser is hampered by its indirect bandgap. Here, we present the integration of a direct bandgap III-V epitaxial layer on top of the SOI waveguide layer by means of a die-to-wafer bonding process in order to realize near-infrared laser emission on and coupled to SOI.

(BEAST), allow surprisingly accurate ex ante predictions of behavior. Unlike the popular models, BEAST does not assume subjective weighting functions or cognitive shortcuts. Rather, it assumes the use of sampling tools and reliance on small samples, in addition to the estimation of the expected values. (PsycINFO Database Record

Abstract High‐brightness GaN‐based emissive microdisplays can be fabricated with different approaches that are listed and described. They consist either of hybridizing a GaN LED array on a CMOS circuit or building a monolithic component on a single substrate. Using the hybridization approach, two types of 10‐μm pixel pitch GaN microdisplay prototypes were developed: (1) directly driven, 300 × 252 pixels and (2) active‐matrix, 873 × 500 pixels. Brightness as high as 1 × 10 6 and 1 × 10 7 cd/m 2 for blue and green arrays, respectively, were reached. GaN‐based emissive microdisplays are suitable for augmented reality systems or head‐up displays, but some challenges remain before they can be put in production.