Publication

Journal papers, Talks, Conference papers, Books, Technical notes

Search

PhD Defense Slides

Méthodes d'optimisation pour la localisation active et passive des cibles

Author: Garcia Nil

Defended in April 2015

Download document

Active and passive localization employing widely distributed sensors is a problem of interest in various fields. In active localization, such as in MIMO radar, transmitters emit signals that are reected by the targets and collected by the receive sensors, whereas, in passive localization the sensors collect the signals emitted by the sources themselves. This dissertation studies optimization methods for high precision active and passive localization. In the case of active localization, multiple transmit elements illuminate the targets from different directions. The signals emitted by the transmitters may differ in power and bandwidth. Such resources are often limited and distributed uniformly among the transmitters. However, previous studies based on the well known Crámer-Rao lower bound have shown that the localization accuracy depends on the locations of the transmitters as well as the individual channel gains between different transmitters, targets and receivers. Thus, it is natural to ask whether localization accuracy may be improved by judiciously allocating such limited resources among the transmitters. Using the Cráamer-Rao lower bound for target localization of multiple targets as a figure of merit, approximate solutions are proposed to the problems of optimal power, optimal bandwidth and optimal joint power and bandwidth allocation. These solutions are computed by minimizing a sequence of convex problems. The quality of these solutions is assessed through extensive numerical simulations and with the help of a lower-bound that certifies their optimality. Simulation results reveal that bandwidth allocation policies have a stronger impact on performance than power. Passive localization of radio frequency sources over multipath channels is a diffcult problem arising in applications such as outdoor or indoor geolocation. Common approaches that combine ad-hoc methods for multipath mitigation with indirect localization relying on intermediary parameters such as time-of-arrivals, time difference of arrivals or received signal strengths, are unsatisfactory. This dissertation models the localization of known waveforms over unknown multipath channels in a sparse framework, and develops a direct approach in which multiple sources are localized jointly, directly from observations obtained at distributed sources. The proposed approach exploits channel properties that enable to distinguish line-of-sight (LOS) from non-LOS signal paths. Theoretical guarantees are established for correct recovery of the sources' locations by atomic norm minimization. A second-order-conebased algorithm is developed to produce the optimal atomic decomposition, and it is shown to produce high accuracy location estimates over complex scenes, in which sources are subject to diverse multipath conditions, including lack of LOS.

Read more

Conference Paper

Change Detection for Optical and Radar Images Using a Bayesian Nonparametric Model Coupled with a Markov Random Field

Authors: Prendes Jorge, Chabert Marie, Pascal Frédéric, Giros Alain and Tourneret Jean-Yves

In Proc. IEEE Int. Conf. Acoust., Speech, and Signal Proc. (ICASSP), Brisbane, Australia, April 19-24, 2015.

Download document

This paper introduces a Bayesian non parametric (BNP) model associated with a Markov random field (MRF) for detecting changes between remote sensing images acquired by homogeneous or heterogeneous sensors. The proposed model is built for an analysis window which takes advantage of the spatial information via an MRF. The model does not require any a priori knowledge about the number of objects contained in the window thanks to the BNP framework. The change detection strategy can be divided into two steps. First, the segmentation of the two images is performed using a region based approach. Second, the joint statistical properties of the objects in the two images allows an appropriate manifold to be defined. This manifold describes the relationships between the different sensor responses to the observed scene and can be learnt from a training unchanged area. It allows us to build a similarity measure between the images that can be used in many applications such as change detection or image registration. Simulation results conducted on synthetic and real optical and synthetic aperture radar (SAR) images show the efficiency of the proposed method for change detection.

Read more

Signal and image processing / Earth observation

Lowpass/Bandpass Signal Reconstruction and Digital Filtering from Nonuniform Samples

Authors: Bonacci David and Lacaze Bernard

In Proc. IEEE Int. Conf. Acoust., Speech, and Signal Proc. (ICASSP), Brisbane, Australia, April 19-24, 2015.

Download document

This paper considers the problem of non uniform sampling in the case of finite energy functions and random processes, not necessarily approaching to zero as time goes to infinity. The proposed method allows to perform exact signal reconstruction, spectral estimation or linear filtering directly from the non-uniform samples. The method can be applied to either lowpass, or bandpass signals.

Read more

Signal and image processing / Other

Enhanced HARQ for Delay Tolerant Services in Mobile Satellite Communications

Authors: Ali Ahmad Rami, Lacan Jérôme, Arnal Fabrice, Gineste Mathieu and Clarac Laurence

In Proc. The Seventh International Conference on Advances in Satellite and Space Communications (SPACOMM), Barcelona, Spain, April 19-24, 2015.

Download document

The objective of our paper is to improve efficiency (in terms of throughput or system capacity) for mobile satellite communications. In this context, we propose an enhanced Hybrid Automatic Repeat reQuest (HARQ) for delay tolerant services. Our proposal uses the estimation of the mutual information. We evaluate the performance of the proposed method for a land mobile satellite channel by means of simulations. Results are compared with those obtained with a classical incremental redundancy (IR) HARQ scheme. The technique we propose, shows a better performance in terms of efficiency while maintaining an acceptable delay for services.

Read more

Digital communications / Space communication systems

A Bayesian Approach for the Joint Estimation of the Multifractality Parameter and Integral Scale Based on the Whittle Approximation

Authors: Combrexelles Sébastien, Wendt Herwig, Abry Patrice, Dobigeon Nicolas, McLaughlin Stephen and Tourneret Jean-Yves

In Proc. IEEE Int. Conf. Acoust., Speech, and Signal Proc. (ICASSP), Brisbane, Australia, April 19-24, 2015.

Download document

Multifractal analysis is a powerful standard signal processing tool. Multifractal models are essentially characterized by two parameters, the so-called multifractality parameter c2 and the integral scale A (the time scale beyond which multifractal properties vanish). Yet, most applications concentrate on estimating c2 while estimating A is mostly overlooked, despite of A potentially conveying important information. Joint estimation of c2 and A is challenging due to the statistical nature of multifractal processes (strong dependence, non-Gaussian), and has barely been considered. The present contribution addresses these limitations and proposes a Bayesian procedure for the joint estimation of (c2;A). Its originality resides, first, in the construction of a generic multivariate model for the statistics of wavelet leaders for multifractal multiplicative cascade processes, and second, in the use of a suitable Whittle approximation for the likelihood associated with the model. The resulting model enables Bayesian estimators for (c2;A) to be computed also for large sample size. Performance is assessed numerically for synthetic multifractal processes and illustrated for wind-tunnel turbulence data. The proposed procedure significantly improves estimation of c2 and yields, for the first time, reliable estimates for A.

Read more

Signal and image processing / Earth observation

Nonlinear Regression Using Smooth Bayesian Estimation

Authors: Halimi Abderrahim, Mailhes Corinne and Tourneret Jean-Yves

In Proc. IEEE Int. Conf. Acoust., Speech, and Signal Proc. (ICASSP), Brisbane, Australia, April 19-24, 2015.

This paper proposes a new Bayesian strategy for the estimation of smooth parameters from nonlinear models. The observed signal is assumed to be corrupted by an independent and non identically (colored) Gaussian distribution. A prior enforcing a smooth temporal evolution of the model parameters is considered. The joint posterior distribution of the unknown parameter vector is then derived. A Gibbs sampler coupled with a Hamiltonian Monte Carlo algorithm is proposed which allows samples distributed according to the posterior of interest to be generated and to estimate the unknown model parameters/hyperparameters. Simulations conducted with synthetic and real satellite altimetric data show the potential of the proposed Bayesian model and the corresponding estimation algorithm for nonlinear regression with smooth estimated parameters.

Read more

Signal and image processing / Earth observation

Unsupervised Unmixing of Hyperspectral Images Accounting for Endmember Variability

Authors: Halimi Abderrahim, Dobigeon Nicolas, Tourneret Jean-Yves and Honeine Paul

In Proc. IEEE Int. Conf. Acoust., Speech, and Signal Proc. (ICASSP), Brisbane, Australia, April 19-24, 2015.

Download document

This paper presents an unsupervised Bayesian algorithm for hyperspectral image unmixing accounting for endmember variability. The pixels are modeled by a linear combination of endmembers weighted by their corresponding abundances. However, the endmembers are assumed random to take into account their variability in the image. An additive noise is also considered in the proposed model generalizing the normal compositional model. The proposed algorithm exploits the whole image to benefit from both spectral and spatial information. It estimates both the mean and the covariance matrix of each endmember in the image. This allows the behavior of each material to be analyzed and its variability to be quantified in the scene. A spatial segmentation is also obtained based on the estimated abundances. In order to estimate the parameters associated with the proposed Bayesian model, we propose to use a Hamiltonian Monte Carlo algorithm. The performance of the resulting unmixing strategy is evaluated via simulations conducted on both synthetic and real data.

Read more

Signal and image processing / Earth observation

Joint Bayesian Deconvolution and Point Spread Function Estimation for Ultrasound Imaging

Authors: Zhao Ningning, Basarab Adrian, Kouamé Denis and Tourneret Jean-Yves

In Proc. Int. Symp. Biomed. Imaging (ISBI'2015), pp. 235-238, New-York, April 16-19, 2015.

This paper addresses the problem of blind deconvolution for ultrasound images within a Bayesian framework. The prior of the unknown ultrasound image to be estimated is assumed to be a product of generalized Gaussian distributions. The point spread function of the system is also assumed to be unknown and is assigned a Gaussian prior distribution. These priors are combined with the likelihood function to build the joint posterior distribution of the image and PSF. However, it is difficult to derive closed-form expressions of the Bayesian estimators associated with this posterior. Thus, this paper proposes to build estimators of the unknown model parameters from samples generated according to the model posterior using a hybrid Gibbs sampler. Simulation results performed on synthetic data allow the performance of the proposed algorithm to be appreciated.

Read more

Signal and image processing / Earth observation

Enhancing Satellite System Throughput Using Adaptive HARQ for Delay Tolerant Services in Mobile Communications

Authors: Ali Ahmad Rami, Lacan Jérôme, Arnal Fabrice, Gineste Mathieu and Clarac Laurence

In Proc. Wireless Telecommunications Symposium WTS 2015, NYC, USA, April 15-17, 2015.

Download document

In this paper we propose the introduction of adaptive hybrid automatic repeat request (HARQ) in the context of mobile satellite communications. HARQ schemes which are commonly used in terrestrial links, can be adapted to improve the throughput for delay tolerant services. The proposed method uses the estimation of the mutual information between the received and the sent symbols, in order to estimate the number of bits necessary to decode the message at next transmission. We evaluate the performance of our method by simulating a land mobile satellite (LMS) channel. We compare our results with the static HARQ scheme, showing that our adaptive retransmission technique has better efficiency while keeping an acceptable delay for services.

Read more

Digital communications / Space communication systems

Robust Kalman Filtering for NLOS Mitigation of GNSS Measurements in Urban Environments

Authors: Kbayer Nabil, Sahmoudi Mohamed, Chaumette Eric and Chapuis Thierry

In Proc. European Navigation Conference (ENC), Bordeaux, France, April 7-10, 2015.

Download document

It is well-known that the Extended Kalman Filer (EKF) is the standard estimation method for positioning with GNSS measurements. However, this filtering method is not optimal when the GNSS measurements become contaminated by non-Gaussian errors including multipath (MP) and non-line-of-sight (NLOS) errors. In this paper, we apply some techniques from robust statistic to make the conventional EKF more resistant to outliers which may be summed up as MP and NLOS signals in urban environments. We study two robust estimators that do not require tuning parameters fixed in advance: the first estimator detect the outliers using a robust statistical test based on the measure of the distance between each innovation sample with respect to the median of all innovations, then assign to them a low weight in the state estimation while keeping nominal weights for good Pseudo-Ranges (PR). The second estimator exploits the difference between two successive innovations to detect jumps related to large errors as MP and NLOS bias and correct their effect via a new recursive weighting technique. Test results using real GPS signal in downtown of Toulouse show that these estimators are simple to implement and capable of detecting multiple outliers in real-time and then improving the positioning accuracy compared to the conventional EKF.

Read more

Signal and image processing / Other

ADDRESS

7 boulevard de la Gare
31500 Toulouse
France

CONTACT


CNES
Thales Alenia Space
Collins Aerospace
Toulouse INP
ISEA-SUPAERO
IPSA
ENAC
IMT Atlantique