Entity

Time filter

Source Type

Le Touquet – Paris-Plage, France

Liu X.,University of Western Australia | Cardoso J.-F.,Ecole Nationale Superieure des Telecommunications | Randall R.B.,University of New South Wales
Mechanical Systems and Signal Processing | Year: 2010

This paper works on joint approximate diagonalization of simplified fourth order cumulant matrices for very fast and large scale blind separation of instantaneous mixing model sources. The JADE algorithm is widely accepted but only limited to small scale separation tasks. The SHIBBS algorithm calculates a fraction of the fourth order cumulant set and avoids eigenmatrix decomposition to reduce calculation cost. However, it was seen to be slower than JADE at the time of its first publication and is hence less known. On the other hand, the SJAD algorithm using the same approach is shown to be very fast. This paper studies the iteration convergence criterion and proposes to use a signal to noise ratio based iteration stopping threshold approach. The improved SHIBBS/SJAD algorithm is very fast, and capable of large scale separation. Experimental separation comparisons between the SHIBBS/SJAD and FastICA are presented. © 2010 Elsevier Ltd. All rights reserved. Source


Larbi S.,National Engineering School of Tunis | Jaidane M.,National Engineering School of Tunis | Moreau N.,Ecole Nationale Superieure des Telecommunications
European Signal Processing Conference | Year: 2015

Digital audiowatermarking can be viewed as a communication system (cf. figure 1) where the information vn is embedded imperceptibly into the digital audio signal xn through appropriate spectral shaping filter Hj. The resulting watermark tn should be robust to standard signal manipulations (i.e. compression, A/D-D/A conversion, resampling,...) on one side, and to intentional pirat attacks on the other side. In particular, all-pass filter attacks introduce a nonminimum phase problem in the watermarking detection scheme of figure 1, which strongly degrades the performances of detection techniques based on mean square criteria. We propose in this paper a solution to the above mentioned problem, combining Wiener deconvolution and blind equalization based on a non quadratic criterion. © 2002 EUSIPCO. Source


Ciblat P.,Ecole Nationale Superieure des Telecommunications | Vandendorpe L.,Catholic University of Louvain
European Signal Processing Conference | Year: 2015

We consider transmission over a frequency-selective channel. We focus on the data-aided joint and directed maximum-likelihood estimation of the carrier frequency offset and of the dispersive channel. The directed estimators correspond to the frequency offset estimator assuming the channel known, as well as the channel estimator assuming the frequency offset known. A comparison of directed and non-directed estimators based on asymptotic (large sample) analysis is addressed and shows that the performance of the joint estimates and the directed estimates is not so different. © 2002 EUSIPCO. Source


Nicolas J.M.,Ecole Nationale Superieure des Telecommunications
European Signal Processing Conference | Year: 2015

Probability Density Functions defined on IR+ can be successfully modeled with the help of 'second kind statistics'. This new approach, proposed in [4], is based on Mellin transform instead of Fourier transform so that classical probability density functions defined on IR+ can be identified with the help of 'second kind moments' and 'second kind cumulants', the analytic expressions of which are oversimple. In this article, we propose to analyse a-stable positive distributions. Indeed we know that classical moments of such distributions are generally not defined. As it is possible to derive their 'second kind moments' and their 'second kind cumulants', the estimation of the parameters of such laws is avalaible with the help of very simple expressions. © 2002 EUSIPCO. Source


Pischella M.,Superieur dElectronique de Paris | Belfiore J.-C.,Ecole Nationale Superieure des Telecommunications
IEEE Transactions on Vehicular Technology | Year: 2010

This paper addresses resource allocation for weighted sum throughput maximization (WSTM) in multicell orthogonal frequency-division multiple-access (OFDMA) networks. Two suboptimal algorithms with polynomial-time complexity are provided. First, we determine a graph-based subcarrier allocation algorithm allowing joint transmission of two interfering links whenever such a transmission fulfills the pairwise WSTM objective. An analytical expression of the optimal power allocation with two interfering cells in single-carrier transmission is obtained, and the capacity region study serves as the basis to build the interference graph. Second, we present a distributed power-control algorithm suitable for any signal-to-interference-plus-noise ratio (SINR) regime. It rejects users and subcarriers with weighted SINR that is too low before operating in the high SINR regime for the remaining users and subcarriers. Both algorithms and their combination are assessed via dynamic simulations, where the weight of each user is proportional to its queue length. They are shown to significantly decrease resource consumption and efficiently balance users' queue lengths. © 2010 IEEE. Source

Discover hidden collaborations