Time filter

Source Type

Gif - sur - Yvette, France

École supérieure d'électricité, commonly known as Supélec, is a French graduate school of engineering awarding the equivalent of a master's degree and Ph.D opportunities. It is one of the most prestigious and selective Grandes Ecoles in France, and a reference in the field of electric energy and information science. With 460 graduates a year, Supélec ranks among the best departments of electrical and computer engineering of the top American or European universities.Founded in 1894 and initially located in the 15th district of Paris, it was moved to Gif-sur-Yvette in 1975. Since then, two more campuses have been established, in Rennes in 1972 and Metz in 1985. It is a member of Top Industrial Managers for Europe network. It is also a member of the CESAER Association and n+i Engineering Studies. Wikipedia.

Tembine H.,Supelec
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics | Year: 2011

In this paper, we study dynamic robust power-allocation games in multiple-input-multiple-output systems under the imperfectness of the channel-state information at the transmitters. Using a robust pseudopotential-game approach, we show the existence of robust solutions in both discrete and continuous action spaces under suitable conditions. Considering the imperfectness in terms of the payoff measurement at the transmitters, we propose a COmbined fully DIstributed Payoff and Strategy Reinforcement Learning (CODIPAS-RL) in which each transmitter learns its payoff function, as well as the associated optimal covariance matrix strategies. Under the heterogeneous CODIPAS-RL, the transmitters can use different learning patterns (heterogeneous learning) and different learning rates. We provide sufficient conditions for the almost-sure convergence of the heterogeneous learning to ordinary differential equations. Extensions of the CODIPAS-RL to It's stochastic differential equations are discussed. © 2011 IEEE. Source

Boyer R.,Supelec
IEEE Transactions on Signal Processing | Year: 2011

To identify a target, the moving noncoherent colocated multiple-input multiple-output (MIMO) radar system takes advantage of multiple antennas in transmission and reception which are close in space. In this paper, we study the estimation performance and the resolution limit for this scheme in which each array geometry is described by the sample-variance of the sensor distribution. So, our analysis encompasses any sensor distributions, including varying intersensors distances or/and lacunar (missing sensors) configuration. As in the space-time MIMO model considered here the radar is moving, the target Doppler frequency cannot be assumed invariant to the target position/angle. The first part of this paper derives and analyzes closed form (nonmatrix) expressions of the deterministic Cramér-Rao lower bound (CRB) for the direction and the velocity of a moving target contaminated by a structured noise (clutter echoes) and a background noise, including the cases of the clutter-free environment and the high signal-to-noise ratio (SNR) regime. The analysis of the proposed expressions of the CRB allows to better understand the characterization of the target. In particular, we prove the coupling between the direction parameter and the velocity of the target is linear with the radar velocity. In the second part, we focus our study on the analytical (closed form) derivation and the analysis of the angular resolution limit (ARL). Based on the resolution of an equation involving the CRB, the ARL can be interpreted as the minimal separation to resolve two closely spaced targets. Consequently, the ARL is a key quantity to evaluate the performance of a radar system. We show that the ARL is in fact quasi-invariant to the movement of the MIMO radar. © 2010 IEEE. Source

Mazenc F.,Supelec | Bernard O.,French Institute for Research in Computer Science and Automation
Automatica | Year: 2011

It is shown that, for any time-invariant exponentially stable linear system with additive disturbances, time-varying exponentially stable interval observers can be constructed. The technique of construction relies on the Jordan canonical form that any real matrix admits and on time-varying changes of coordinates for elementary Jordan blocks which lead to cooperative linear systems. The approach is applied to detectable linear systems. © 2010 Elsevier Ltd. All rights reserved. Source

Molinie P.,Supelec
IEEE Transactions on Plasma Science | Year: 2012

We present here a wide review of the existing models accounting for surface potential decay, developed in various areas of electrostatics. Several polarization or transport processes may be involved. Though most of the models initially stemmed from electrostatics and semiconductor physics, around the notion of mobility, experiments on polymers often require " thermodynamic" models, describing progressive charge detrapping. Different physical processes likely to be involved in the potential decay (dipolar relaxation, induced conductivity, dispersive transport, slow detrapping) can lead in disordered materials to the same time response, the challenge being to design inventive procedures to distinguish them. A particular attention will be given to the application of these models to polyimide and fluorinated ethylene propylene. The specific case of radiation-induced conductivity in these materials will also be examined in detail. © 2011 IEEE. Source

Hoydis J.,Alcatel - Lucent | Ten Brink S.,Alcatel - Lucent | Debbah M.,Supelec
IEEE Journal on Selected Areas in Communications | Year: 2013

We consider the uplink (UL) and downlink (DL) of non-cooperative multi-cellular time-division duplexing (TDD) systems, assuming that the number N of antennas per base station (BS) and the number K of user terminals (UTs) per cell are large. Our system model accounts for channel estimation, pilot contamination, and an arbitrary path loss and antenna correlation for each link. We derive approximations of achievable rates with several linear precoders and detectors which are proven to be asymptotically tight, but accurate for realistic system dimensions, as shown by simulations. It is known from previous work assuming uncorrelated channels, that as N→∞ while K is fixed, the system performance is limited by pilot contamination, the simplest precoders/detectors, i.e., eigenbeamforming (BF) and matched filter (MF), are optimal, and the transmit power can be made arbitrarily small. We analyze to which extent these conclusions hold in the more realistic setting where N is not extremely large compared to K. In particular, we derive how many antennas per UT are needed to achieve η% of the ultimate performance limit with infinitely many antennas and how many more antennas are needed with MF and BF to achieve the performance of minimum mean-square error (MMSE) detection and regularized zero-forcing (RZF), respectively. © 2012 IEEE. Source

Discover hidden collaborations