Search

The purpose of this paper is to share our positive experience about the collaboration launched a few years ago between UTP (Panama) and Bordeaux INP (France) in the field of signal processing. This collaboration involves research and education activities. This has led to numerous internships of French students in Panama, mobilities of researchers, common research papers, and the 1st double diploma signed between France and a country of Central America. Thus, this paper presents the various aspects of the collaboration.

Micro-Doppler induced by mechanical vibrating or rotating structures in a radar target is possibly useful for its detection, classification and recognition. In a previous work, pseudo-Zernike moments (PZMs) were used as micro-Doppler features for classification. Despite of their promising classification rates, the choice of PZMs is debatable because other types of moments exist. In this paper, our purpose is to compare various kinds of micro-Doppler features such as Zernike moments, PZMs, orthogonal Mellin-Fourier moments, Legendre moments and Krawtchouk moments in order to evaluate which moments are the most relevant in terms of reconstruction ability, computational cost and micro-Doppler classification rate. Advantages and drawbacks of each family of moments are also given. Through the simulations we carried out, when the signal is disturbed by an additive white noise and the signal-to-noise ratio is low, the use of Krawtchouk moments as micro-Doppler features turns out to be the best compromise.

In some multi-object tracking scenarios such as convoys or road constrained motions, it can be advantageous to track groups of targets sharing common motion characteristics, even if they are not necessarily close to each other. The objective is twofold: reducing the computational cost while increasing the accuracies of the individual trajectory estimates. In a previous communication, we introduced a generic model based on hierarchical random finite sets (RFSs) to represent these types of multigroup multi-target scenarios. A first RFS is used to represent the multi-group state: the number of groups, their common motion characteristics and their target compositions are assumed to be random variables. Then, for each group, a second layer of RFSs represents the multi-target state assuming that the number of targets inside a group and their trajectories are also random variables. The main contribution of this paper is to derive a filter dedicated to the state estimation of hierarchical RFSs from sequential sensor measurements. The proposed solution is based on the labeled multi-Bernoulli filter to estimate the group characteristics, which interacts with a bank of probability hypothesis density filters to address the multi-target layer.

Jeffrey’s divergence (JD) is used in many applications, from change detection to classification. Several studies were done on the JD between ergodic wide-sense stationary autoregressive and moving average (ARMA) processes. It was shown that the derivate of the JD between the probability density functions of k consecutive samples of two ARMA processes tends to the so-called asymptotic JD increment. This latter is enough to compare the processes and amounts to calculating the power of the first process filtered by the inverse filter associated with the second process and conversely. In this paper, our purpose is to study if this result can be extended to ARFIMA processes. As a first step, a special case, namely the JD between wide-sense stationary fractionally integrated white noises, is addressed. The influences of the process parameters on the asymptotic JD increment are analyzed. Our investigations validate the inverse filtering interpretation of the JD.

In this paper, single-target tracking using radar measurements is addressed. Recently, algorithms based on Bernoulli random finite sets have proved efficient in a cluttered environment. However, in Bayesian approaches, the choice of the motion model impacts the trajectory estimation accuracy. To select an appropriate set of motion models, a joint tracking and classification (JTC) algorithm can be used. The principle is to consider different target classes depending on their maneuvrability, each of them being associated to a set of motion models. In this context, additional information such as a target length extent measurement can improve both classification and trajectory estimation. Therefore, we propose a multiple-model Bernoulli filter to perform JTC. To jointly estimate the trajectory and the target length which is constant, a Rao-Blackwellized approach is considered. Another contribution is that a bank of probabilistic data association filters is run instead of Kalman filters to account for false detections.

The autoregressive models (AR) and moving-average models (MA) are regularly used in signal processing. Previous works have been done on dissimilarity measures between AR models by using a Riemannian distance, the Jeffrey's divergence (JD) and the spectral distances such as the Itakura-Saito divergence. In this paper, we compare the Rao distance and the JD for MA models and more particularly in the case of 1st-order MA models for which an analytical expression of the inverse of the covariance matrix is available. More particularly, we analyze the advantages of the Rao distance use. Secondly, the simulation part compares both dissimilarity measures depending on the MA parameters but also on the number of data available.

For a high-resolution radar, an extended target is characterized by a few main scatterers spread over several range gates not necessarily consecutive. The joint detections and localizations of these scatterers are of particular interest in order to estimate the range profile and identify the target. In this paper, we study a detector based on the Generalized Likelihood Ratio Test considering the unknown locations of the scatterers. Given the number of scatterers and using order statistics results, we derive approximations of the probability of false alarm and the probability of detection and localization which can be a relevant measure of performance in this context. Finally, a comparative study is carried out with other existing detectors.

When using Bayesian estimation techniques for target tracking, the algorithm accuracy is induced by the choice of the system evolution model. Information on the type of target and its maneuver capability can then be helpful to choose relevant motion models. Joint tracking and classification (JTC) methods based on target features have thus been introduced. Among them, we recently proposed to take into account the target extent measurements for single-target tracking. In this paper, we extend this work to multi-target tracking (MTT) by using probability hypothesis density (PHD) filters. More precisely, assuming that each target class is characterized by its own kinematic-model set, a multiple-model (MM) PHD filter is used for each class. State estimates from each class are then combined by using class probabilities. Finally, the proposed approach, namely a multiclass MM-GMPHD, is applied to maritime-target tracking and simulation results show the relevance of the proposed approach regarding the tracking of various types of targets.

This work aims at improving the power amplifier (PA) efficiency in uplink OFDM-based cognitive radio (CR) communications. Unlike the traditional approaches, we suggest transmitting a non-linearily ampliied signal without any il-tering and addressing the OFDM sample estimation from the distorted signal at the receiver. The proposed post-distortion and detection technique is based on a Volterra model for the PA and the channel. As the transmission can switch from one sub-band to another, the CR-PA behavior varies over time and the Volterra kernels can be constant or suddenly change. Therefore, an interactive multiple model (IMM) combining extended Kalman filters is considered. The transition probability matrix, which plays a key role in the IMM, is also sequentially estimated. The resulting uplink system has various advantages: it learns from the observations and a part of the computational load is exported to the receiver, which is not battery driven unlike the mobile terminal.

Dirichlet process (DP) mixtures were recently introduced to deal with switching linear dynamical models (SLDM). They assume the system can switch between an a priori infinite number of state-space representations (SSR) whose parameters are on-line inferred. The estimation problem can thus be of high dimension when the SSR matrices are unknown. Nevertheless, in many applications, the SSRs can be categorized in different classes. In each class, the SSRs are characterized by a known functional form but differ by a reduced set of unknown hyperparameters. To use this information, we thus propose a new hierarchical model for the SLDM wherein a discrete variable indicates the SSR class. Conditionally to this class, the distributions of the hyperparameters are modeled by DPs. The estimation problem is solved by using a Rao-Blackwellized particle filter. Simulation results show that our model outperforms existing methods in the field of target tracking.

International audience; In the field of cognitive radio (CR), radio frequency (RF) transceivers must be efficient to save the terminal battery autonomy. Therefore, when designing the CR power amplifier (CR-PA), an obvious objective is to optimize efficiency over a large bandwidth. As a consequence, the CR-PA operates in its non-linear region and then frequency-dependent distortions are generated. This issue is all the more critical as one deals with high peak-to-average power ratio (PAPR) like those of OFDM signals. In this paper, we develop a digital post-distortion and detection technique in order to compensate the non-linearities generated by the CR-PA. It is based on a dynamic Volterra model to take into account the non-linear behavior of the CR-PA. The key feature of the proposed technique is the joint estimations of the model parameters and the CR-PA input samples. For this reason, an extended Kalman filter (EKF) is considered. However, as the model parameters can vary over time, several EKFs are combined by means of an interacting multiple model (IMM) algorithm. Simulation results confirm the relevance of the proposed postdistortion and detection technique.

This paper presents a spectrum-based method to estimate the integral non-linearity (INL) of an analog-to-digital converter. It relies on a previous work published in the literature that consists in expanding in Fourier series the analytical expression of a converter distorted signal output. The INL estimation process is then reduced to three operations: a discrete Fourier transform (DFT), a matrix inversion and an inverse DFT. In comparison with the initial approach, the algorithm proposed in this article reduces by two the set of parameters required to obtain the INL. This new method is validated on experimental data from the 12-bit ADC12D800RF converter run at 800MHz.

One of the most important challenges when applying multiple-model approaches is model-set design. However, to our knowledge, there are not general rules to choose models. For this purpose, we recently proposed an approach based on the Jeffrey's divergence to characterize dissimilarities between two state models. In this paper, our contribution is to use the Jeffrey's divergence to classify at least two models into subsets. Our approach consists in creating a dissimilarity matrix composed of Jeffrey's divergences between model pairs. Then, we transform this matrix to get a correlation-like matrix and an eigenvalue decomposition is computed. We propose an interpretation of the predominant eigenvalues and use it to deduce the number of model subsets and their cardinals. Finally, a classification algorithm can be considered to determine which models belong to which subsets. Among the applications, we focus on target tracking.

In current airborne radars, detection and identification are done with two distinct waveforms. However, while the radar switches from one function to another, the target scene can change. In this paper, we propose a hybrid waveform combining intra and interpulse phase codes. This waveform is first processed in a low-resolution “channel”. Then, if a target is detected, the received signal is reprocessed in a high-resolution “channel”. Given the above scenario and assuming that the received data are disturbed by a colored Gaussian clutter, we address the optimization of the phase codes by searching the Pareto front of a multi-objective optimization problem. Finally, we aim at analyzing the sensitivity of the Pareto front to the clutter properties. For this purpose, we study whether a first-order autoregressive (AR) modelling for the clutter in the high-resolution case is relevant or not.

This paper deals with a system combining orthogonal frequency division multiplexing (OFDM), interleave-division multiple access (IDMA) and space time block coding (STBC). Less investigated in the literature than multicarrier code division multiple access or OFDMA, OFDM-IDMA can be viewed as a promising technique to solve the problem of bandwidth scarcity. With a specific frame organization to estimate the carrier frequency offsets (CFO) and the channels, we suggest a new receiver architecture. Indeed, based on a modification of the standard IDMA receiver where a CFO correction is no longer necessary unlike common OFDM approaches, we show that the transmitted bits can be still recovered. When considering a STBC-OFDM-IDMA uplink system over different Rayleigh channels, simulation results show the efficiency of the proposed algorithm. These results confirm the utility of this technique as a mid-term solution to solve the constant increase of data rate and user density per cell in cellular systems.

In radar processing, inverse synthetic aperture radar (ISAR) takes advantage of the target rotational motion to provide a 2-D image of the target. The “quality” of this image must be sufficient to be used for target recognition. However, among the main problems to be addressed, the translation motion which is another component of the target motion must be foremost compensated. If this is not the case, a blurred image is obtained. In this paper, we propose a new approach for translation-motion compensation based on the target radial velocity estimation. This method consists in characterizing the radar echoes by several features (mean and wavelet coefficients for instance) and in storing them in sequences which represent the parameter evolution over time. Spectrum analysis methods are then used to get Doppler frequency estimates. To make the estimation robust against additive noise and sea clutter, several estimates are combined to deduce the target radial velocity estimation. Tests on experimental data confirm the validity of the approach.

In ISAR processing, ship classification is usually based on a range-Doppler (RD) image. The RD image is obtained by combining range alignment, phase compensation and short-time Fourier transform. Nevertheless, this leads to sophisticated approaches to weaken unwanted phenomena such as blurred images. In this paper, an alternative approach is proposed based on the time-frequency analysis. In that case, the Wigner-Ville transform provides geometrical structured events which correspond to the scatterers and unstructured ones which are associated to the artefacts and cross-terms. Taking advantage of these geometrical features, we first suggest computing a time-frequency local gradient analysis. This variational analysis provides a confidence map of the time-frequency distribution we can use to remove artefacts. Then, the practitioner can exploit the corresponding time-frequency images to choose the more appropriate instantaneous RD image. Simulation results confirm the effectiveness of this approach.

In ISAR processing, post-processing of the range Doppler image is useful to help the practitioner for ship recognition. Among the image post-processing tools, interpolation methods can be of interest especially when zooming. In this paper, we study the relevance of the Holderian regularity-based interpolation. In that case, interpolating consists in adding a new scale in the wavelet transform and the new wavelet coefficients can be estimated from others. In the original method, initially proposed by two of the authors, the image is first interpolated along the rows and then along the columns. Concerning the diagonal pixels, they are estimated as the mean of the adjacent original and interpolated pixels. Here, we propose a variant where the diagonal pixels are estimated by taking into account the local orientation of the image. It has the advantage of conserving local regularity on all interpolated pixels of the image. A comparative study on synthetic data and real range-Doppler images is then carried out with alternative interpolation techniques such as the linear interpolation, the bicubic one, the nearest neighbour interpolation, etc. The simulation results confirm the effectiveness of the approach.

A great deal of interest has been paid to the estimation of time-varying autoregressive (TVAR) parameters. However, when the observations are disturbed by an additive white measurement noise, using standard least squares methods leads to a weight-estimation bias. In this paper, we propose to jointly estimate the TVAR parameters and the measurement-noise variance from noisy observations by means of a generalized eigenvalue decomposition. It extends to the TVAR case an off-line method that was initially proposed for AR parameter estimation from noisy observations. A comparative study is then carried out with existing methods such as the recursive errors-in-variable approach and Kalman based algorithms.

In Cognitive Radio (CR) systems, spectrum sensing plays a key role to determine the free frequency bands. However, when the primary-user (PU) signal spectrum exhibits localized fading, PU detection cannot be guaranteed. In addition, as the CR may use the PU faded frequencies, the PU spectrum can be disturbed by a narrow-band interference (NBI) and synchronization algorithms used for the PU carrier frequency offset (CFO) estimation suffer degradations. In this paper, we propose a new scheme that jointly allows the CR-NBI to be detected and the PU-CFOs and the channels to be estimated in an orthogonal frequency division multiple access (OFDMA) system. It combines a sigma point Kalman filter and a test aiming at detecting a variation of the measurement-noise covariance matrix. Simulation results confirm that the proposed algorithm can accurately detect the CR-NBI and estimate the PU-CFOs.

Space time adaptive processing (STAP) is useful in radar processing to detect a target by filtering the clutter and the additive thermal noise. A derived version based on a multichannel autoregressive (M-AR) model of the clutter has the advantage of reducing the computational cost. Nevertheless, the estimation of the AR matrix parameters is a key issue because the clutter is not Gaussian in real cases. When dealing with an off-line solution, the multichannel least squares method (MLS) can be considered, but the estimation of the disturbance covariance matrix is required. In this paper, we suggest using the so-called fixed point method since it has “matrix- and texture-constant false alarm rate” property (matrix-CFAR and texture-CFAR) and it provides an unbiased and consistent estimate in a non-Gaussian case. A comparative study is then carried out between off-line M-AR based STAP methods and it points out the relevance of the solution we propose.

This paper deals with the issue of correlated noises in GPS navigation. GPS is based on the measure of the propagation delays of satellite signals. Therefore, additional delays induced when traveling through the ionosphere or the troposphere degrade GPS accuracy. These error sources are correlated, both spatially and temporally. Thus, when using an extended Kalman filter as navigation algorithm, these correlations should be taken into account to ensure that an optimal solution is obtained in terms of mean square error. Our contribution is to study, in this context, the relevance of an alternative approach well-known in the field of control engineering: the H ∞ filter. Also based on a state representation, this technique has the advantage of relaxing the constraints on the measurement and state noises. A comparative study with a standard extended Kalman filter and a colored extended Kalman filter is carried out to illustrate which of the above-mentioned approaches achieves the better compromise between accuracy and computational complexity.

In civil aviation applications, Satellite failures yield unacceptable positioning errors when using the Global Positioning System (GPS). To ensure the user security, the navigation system has to fulfill stringent performance requirements. Thus, detecting and excluding the faulty GPSmeasurements is necessary prior to estimating the mobile location. Classical fault detection algorithms based on Kalman filters (KF) are sensitive to the choice of an appropirate motion model for the mobile. To overcome this difficulty, we propose in this paper a new fault detection algorithm wherein the KF are replaced by multiple model algorithms. In this way, both the false alarm rate and the positioning mean square error are shown to be decreased.

AR modeling is used in a wide range of applications from speech processing to Rayleigh fading channel simulation. When the observations are disturbed by an additive white noise, the standard Least Squares estimation of the AR parameters is biased. Some authors of this paper recently reformulated this problem as an errors-in-variables (EIV) issue and proposed an off-line solution, which outperforms other existing methods. Nevertheless, its computational cost may be high. In this paper, we present a blind recursive EIV method that can be implemented for real-time applications. It has the advantage of converging faster than the noise-compensated LMS based solutions. In addition, unlike EKF or Sigma Point Kalman filter, it does not require a priori knowledge such as the variances of the driving process and the additive noise. The approach is first tested with synthetic data; then, its relevance is illustrated in the field of radar sea clutter rejection.

In 1952 Hodgkin and Huxley introduced the voltage--clamp technique to extract the parameters of the ionic channel model of a neuron. Although this method is widely used today, it has a lot of disadvantages. In this paper, we propose an alternative approach to the estimation method of the voltage--clamp technique using metaheuristics such as Simulated Annealing, Genetic Algorithms and Differential Evolution. This method avoids approximations of the original technique by simultaneously estimating all the parameters of a single ionic channel with a single fitness function. To compare the different methods, we apply them on measurements from a neuromimetic integrated circuit. This circuit, due to its analog behavior, provides us noisy data like a biological system. Therefore we can validate the efficiency of our method on experimental-like data.

In this paper, we propose a new Rayleigh channel simulator. Modeling the channel by an AR process leads to numerical problems due to the bandlimitation of the theoretical power density spectrum (PSD) of a Rayleigh channel. Therefore, we suggest modeling the channel by a low-pass filtered version of the so-called stochastic sinusoidal process. It consists of sinusoids in quadrature with random magnitudes modeled as AR processes. To estimate the AR parameters of the amplitudes, we take advantage of the asymptotic behavior of the first-kind zero-order Bessel function. We show that unlike an AR channel modeling, this simulator has the advantage of exhibiting the PSD peaks at the maximum Doppler frequency, for any AR process order.

Among parametric methods for speech enhancement, one consists in combining an autoregressive model for speech and a Kalman filter. This filtering is optimal in the H/sub 2/ sense providing the initial state vector, the input and the observation vectors in the state space representation of the system are independent, white and Gaussian. However, these assumptions do not necessarily hold when processing speech. In this paper, we propose to investigate an alternative approach, which is based on H/sub /spl infin// filtering and hence does not depend on these restrictive assumptions. In that setting, the purpose is to minimize the worst possible effects of the noises and system uncertainties on the estimation error. A comparative study between Kalman and H/sub /spl infin// filtering is carried out, when the additive colored noise can be modeled by a moving average (MA) process.

This paper deals with the identification of time-varying frequency-flat Rayleigh fading channels disturbed by an additive white Gaussian noise, using a training sequence based approach. When the channel is modeled by an Autoregressive (AR) process, it can be estimated by using a Kalman filter. However, this solution requires the preliminary unbiased estimations of the AR parameters and the variances of both the additive noise and the driving process in the state space representation of the system. Instead of using the existing noise compensated approaches which usually require a long observation window and do not necessarily provide reliable estimates when the signal to noise ratio is low, we propose an alternative approach using recent results developed for the Errors-In-Variables (EIV) issue. This method consists in estimating the kernel of specific autocorrelation matrices and has the advantage of providing both the noise variances and the channel AR parameters. Moreover, the maximum Doppler frequency can be also deduced.

The paper deals with the local sensitivity of the discrete-time infinite-horizon H/sup /spl infin// filtering problem. A new nonlinear perturbation bound is derived for the solution of the related matrix Riccati equation.

Kalman algorithms have been widely applied, for instance in single-channel speech enhancement. However, when carrying out Kalman smoothing, computational cost and data storage requirements are two specific problems. A dual-filter-based smoother is proposed and used in the framework of speech enhancement. Our approach comprises a forward-in-time Kalman filter and a backward-in-time Kalman filter. Both filters are based on their respective forward-in-time linear prediction (LP) model and backward-in-time LP model. This method does not require as large a storage space as a standard Kalman smoother does. The algorithm is evaluated by considering a speech signal embedded in a white Gaussian noise. Simulation results show that the proposed algorithm provides a higher improvement of signal-to-noise ratio (SNR) than Kalman filtering.

The paper deals with the local sensitivity of the discrete-time H/sub /spl infin// filtering problem. A condition number based perturbation bound is obtained for the solution of the Riccati equation which determines the sensitivity of the problem.