Searcharxiv⌕ Search

arXiv subjects

Julien Sylvestre

Publications and source records attributed to Julien Sylvestre.

13 recordsLinked to original sources

Microfabricated Neuroaccelerometer: Integrating Sensing and Reservoir Computing in MEMS

This study presents the design, fabrication, and test of a micro accelerometer with intrinsic processing capabilities, that integrates the functions of sensing and computing in the same MEMS. The device consists of an inertial mass electrostatically coupled to an oscillating beam through a gap of 8 μm. The motion of the inertial mass modulates an AC electrostatic field that drives the beam in its non-linear regime. This non-linearity is used to implement machine learning in the mechanical domain, using reservoir computing with delayed feedback to process the acceleration information provided by the inertial mass. The device is microfabricated on a silicon-on-insulator substrate using conventional MEMS processes. Dynamic characterization showed good accelerometer functionalities, with an inertial mass sensitivity on the order of 100 mV/g from 250 to 1300 Hz and a natural frequency of 1.7 kHz. In order to test the device computing capabilities, two different machine learning benchmarks were implemented, with the inputs fed to the device as accelerations. The neuromorphic MEMS accelerometer was able to accurately emulate non-linear autoregressive moving average models and compute the parity of random bit streams. These results were obtained in a test system with a non-trivial transfer function, showing a robustness that is well-suited to anticipated applications.

cs.ET↗

FPGA-based Multi-Chip Module for High-Performance Computing

Current integration, architectural design and manufacturing technologies are not suited for the computing density and power efficiency requested by Exascale computing. New approaches in hardware architecture are thus needed to overcome the technological barriers preventing the transition to the Exascale era. In that scope, we report successful fabrication of first ExaNoDe's MCM prototypes dedicated to Exascale computing applications. Each MCM was composed of 2 Xilinx Zynq Ultrascale+ MPSoC, assembled on advanced 68.5 mm x 55 mm laminate substrates specifically designed and fabricated for the project. Acoustic microscopy, x-ray, cross-section and Thermo-Moire investigations revealed no voids, shorts, delamination, cracks or warpage issues. Two MCMs were mounted on a daughter board by FORTH for testing purposes. The DDR memories on the 4 SODIMMs of the daughter board were successfully tested by running extensive Xilinx memory tests with clock frequencies of 1866 MHz and 2133 MHz. All 4 FPGAs were programmed with the Xilinx integrated bit error ratio test (IBERT) tailored for this board for links testing. All intra-board high-speed links between all FPGAs were stable at 10 Gbps, even under the more demanding 31-bit PRBS (Pseudorandom Binary Sequence) tests.

cs.AR↗

Dynamical response of a radiative thermal transistor based on suspended insulator-metal transition membranes

We investigate the dynamical control of the heat flux exchanged in near-field regime between a membrane made with a phase-change material and a substrate when the temperature of the membrane is tuned around its critical value. We show that in interaction with an external source of thermal radiation, this system is multistable and behaves as a thermal transistor, being able to dynamically modulate and even amplify super-Planckian heat fluxes. This behavior could be used to dynamically control heat fluxes exchanged at the nanoscale in systems out of thermal equilibrium and to process thermal information employing suspended membranes.

cond-mat.mes-hall↗

Computing with Networks of Nonlinear Mechanical Oscillators

As it is getting increasingly difficult to achieve gains in the density and power efficiency of microelectronic computing devices because of lithographic techniques reaching fundamental physical limits, new approaches are required to maximize the benefits of distributed sensors, micro-robots or smart materials. Biologically-inspired devices, such as artificial neural networks, can process information with a high level of parallelism to efficiently solve difficult problems, even when implemented using conventional microelectronic technologies. We describe a mechanical device, which operates in a manner similar to artificial neural networks, to solve efficiently two difficult benchmark problems (computing the parity of a bit stream, and classifying spoken words). The device consists in a network of masses coupled by linear springs and attached to a substrate by non-linear springs, thus forming a network of anharmonic oscillators. As the masses can directly couple to forces applied on the device, this approach combines sensing and computing functions in a single power-efficient device with compact dimensions.

cs.ET↗

Simulations of laser locking to a LISA arm

We present detailed numerical simulations of a laser phase stabilization scheme for LISA, where both lasers emitting along one arm are locked to each other. Including the standard secondary noises and spacecraft motions that approximately mimic LISA's orbit, we verify that very stable laser phases can be obtained, and that time delay interferometry can be used to remove the laser phase noise from measurements of gravitational wave strains. Most importantly, we show that this locking scheme can provide significant simplifications over LISA's baseline design in the implementation of time delay interferometry.

gr-qc↗

Noise characterization for LISA

We consider the general problem of estimating the inflight LISA noise power spectra and cross-spectra, which are needed for detecting and estimating the gravitational wave signals present in the LISA data. For the LISA baseline design and in the long wavelength limit, we bound the error on all spectrum estimators that rely on the use of the fully symmetric Sagnac combination ($ζ$). This procedure avoids biases in the estimation that would otherwise be introduced by the presence of a strong galactic background in the LISA data. We specialize our discussion to the detection and study of the galactic white dwarf-white dwarf binary stochastic signal.

gr-qc↗

Optimal generalization of power filters for gravitational wave bursts, from single to multiple detectors

Searches for gravitational wave signals which do not have a precise model describing the shape of their waveforms are often performed using power detectors based on a quadratic form of the data. A new, optimal method of generalizing these power detectors so that they operate coherently over a network of interferometers is presented. Such a mode of operation is useful in obtaining better detection efficiencies, and better estimates of the position of the source of the gravitational wave signal. Numerical simulations based on a realistic, computationally efficient hierarchical implementation of the method are used to characterize its efficiency, for detection and for position estimation. The method is shown to be more efficient at detecting signals than an incoherent approach based on coincidences between lists of events. It is also shown to be capable of locating the position of the source.

gr-qc↗

Prospects for the detection of electromagnetic counterparts to gravitational wave events

Various models for electromagnetic emissions correlated with the gravitational wave signals expected to be detectable by the current and planned gravitational wave detectors are studied. The position error on the location of a gravitational wave source is estimated, and is used to show that it could be possible to observe the electromagnetic counterparts to neutron star-neutron star or neutron star-black hole binary coalescences detected with the Advanced LIGO and the Virgo detectors.

astro-ph↗

Implementation of Time-Delay Interferometry for LISA

We discuss the baseline optical configuration for the Laser Interferometer Space Antenna (LISA) mission, in which the lasers are not free-running, but rather one of them is used as the main frequency reference generator (the {\it master}) and the remaining five as {\it slaves}, these being phase-locked to the master (the {\it master-slave configuration}). Under the condition that the frequency fluctuations due to the optical transponders can be made negligible with respect to the secondary LISA noise sources (mainly proof-mass and shot noises), we show that the entire space of interferometric combinations LISA can generate when operated with six independent lasers (the {\it one-way method}) can also be constructed with the {\it master-slave} system design. The corresponding hardware trade-off analysis for these two optical designs is presented, which indicates that the two sets of systems needed for implementing the {\it one-way method}, and the {\it master-slave configuration}, are essentially identical. Either operational mode could therefore be implemented without major implications on the hardware configuration. We then.......

gr-qc↗

Time-frequency detection algorithm for gravitational wave bursts

An efficient algorithm is presented for the identification of short bursts of gravitational radiation in the data from broad-band interferometric detectors. The algorithm consists of three steps: pixels of the time-frequency representation of the data that have power above a fixed threshold are first identified. Clusters of such pixels that conform to a set of rules on their size and their proximity to other clusters are formed, and a final threshold is applied on the power integrated over all pixels in such clusters. Formal arguments are given to support the conjecture that this algorithm is very efficient for a wide class of signals. A precise model for the false alarm rate of this algorithm is presented, and it is shown using a number of representative numerical simulations to be accurate at the 1% level for most values of the parameters, with maximal error around 10%.

gr-qc↗

A nonlinear detection algorithm for periodic signals in gravitational wave detectors

We present an algorithm for the detection of periodic sources of gravitational waves with interferometric detectors that is based on a special symmetry of the problem: the contributions to the phase modulation of the signal from the earth rotation are exactly equal and opposite at any two instants of time separated by half a sidereal day; the corresponding is true for the contributions from the earth orbital motion for half a sidereal year, assuming a circular orbit. The addition of phases through multiplications of the shifted time series gives a demodulated signal; specific attention is given to the reduction of noise mixing resulting from these multiplications. We discuss the statistics of this algorithm for all-sky searches (which include a parameterization of the source spin-down), in particular its optimal sensitivity as a function of required computational power. Two specific examples of all-sky searches (broad-band and narrow-band) are explored numerically, and their performances are compared with the stack-slide technique (P. R. Brady, T. Creighton, Phys. Rev. D, 61, 082001).

gr-qc↗

Turbulent Bipolar Outflows in Young Stellar Objects: Multifractal Universality Classes and Generalized Scaling

We present a study of the scaling properties of bipolar outflows, generic to the circumstellar environment of young stellar objects, and which are generally believed to be magnetohydrodynamically driven. Our work consists of a multifractal analysis of near-infrared light scattered by dust grains in the lobes of seven bipolar nebulae, namely GGD 18, V380 Orionis, LkHα101/NGC 1579, LkHα233, PV Cephei, V645 Cygni (GL 2789), and V633 Cassiopeiae (LkHα198). We argue that these objects pertain to a multifractal universality class, with the corresponding universal parameters averaged over the ensemble yielding alpha= 1.96 +/- 0.02, C_1=0.04 +/- 0.02, and H=0.7 +/- 0.2; these results might suggest that the dynamics of the observed objects are similar. We then proceed to investigate the existence of anisotropy in the scaling of GGD 18 using the Generalized Scale Invariance (GSI) formalism. It is found that differential rotation and stratification are involved in the outflow mechanism. The stratification is believed to be dynamical in nature i.e., gravity is apparently not the dominant stratifying force.

astro-ph↗