Search

This work deals with adaptive predictive deconvolution of non-stationary channels. In particular, we investigate the use of a cascade of linear predictors in the recovering of sparse and antisparse original signals. To do so, we first discuss the behavior of the Lp Prediction Error Filter (PEF), with p different of 2, showing that it has a superior ability to deal with non-minimum phase channels in comparison with the classical L2 PEF, although it still presents intrinsic limitations due to its direct linear structure. The cascade structure emerges as a possible solution to circumvent this issue. We apply the proposed cascade structure in the deconvolution of non-stationary channels, with minimum-, maximum- , mixed- and variable-phase response, and also noise scenarios. From the simulation results we observed that, besides the duality relation between the Lp norms, they present different algorithmic behavior: the L1 norm attains a fast convergence, enhancing the cascade tracking capacity, but is more sensible to noise. The L4 norm, on the other hand, is more robust to noise, but presents slower convergence and tracking capability.

Purpose: to characterize the distribution and mechanisms involved in ceftolozane/tazobactam (C/T) resistance in Pseudomonas aeruginosa isolates recovered from intensive care units (ICUs) in Portugal as part of the STEP surveillance study. Materials and methods: a total of 226 P. aeruginosa isolates were collected from patients with low respiratory tract infections (LRTI) admitted to ICUs between June 2017 and July 2018. Susceptibility to antimicrobials including the recent C/T combination was determined by EUCAST-criteria. Whole genome sequencing (WGS) was performed in a subset of 17 isolates. Results: multidrug resistant (MDR) and extremely drug resistant (XDR) phenotypes accounted for 20.4% and 25.7% of cases, respectively. C/T showed the highest susceptibility rate in both MDR (100%) and XDR (93.1%) isolates, followed by amikacin (97.8% MDR and 79.3% XDR). blaKPC-3 (n = 2) and blaGES-13 (n = 1) carbapenemase genes were detected in 3 of the 17 sequenced isolates, but only the GES-13-producing isolate displayed resistance to C/T. Additionally, the C/T-resistant phenotype was also found in two non-carbapenemase producers that carried known ceftolozane/tazobactam resistance-associated mutations in the PBP3 gene. Conclusions: C/T was highly active against MDR/XDR-P. aeruginosa isolates causing LRTI in ICUs. Moreover, beyond carbapenemase-encoding genes, mutations in chromosomal PBP-encoding genes might also be involved in ceftolozane/tazobactam resistance in Portugal.

This letter introduces the concept of antisparse Blind Source Separation (BSS), proposing a suitable criterion based on the $\ell_\infty$ norm to explore the antisparsity feature. The effectiveness of the criterion is theoretically demonstrated and it is also evaluated by computational simulations, which consider up to ten distinct sources with different correlation levels. Moreover, we simulated a scenario in wireless communication with binary sources, comparing our approach to the Constant Modulus algorithm. Both the theoretical and the simulation results highlight the potentiality of using antisparsity as a prior in BSS.

CrpP enzymes have been recently described as a novel ciprofloxacin-resistance mechanism. We investigated by whole genome sequencing the presence of crpP-genes and other mechanisms involved in quinolone resistance in MDR/XDR-Pseudomonas aeruginosa isolates (n = 55) with both ceftolozane-tazobactam susceptible or resistant profiles recovered from intensive care unit patients during the STEP (Portugal) and SUPERIOR (Spain) surveillance studies. Ciprofloxacin resistance was associated with mutations in the gyrA and parC genes. Additionally, plasmid-mediated genes (qnrS2 and aac(6′)-Ib-cr) were eventually detected. Ten chromosomal crpP-like genes contained in related pathogenicity genomic islands and 6 different CrpP (CrpP1-CrpP6) proteins were found in 65% (36/55) of the isolates. Dissemination of CrpP variants was observed among non-related clones of both countries, including the CC175 (Spain) high-risk clone and CC348 (Portugal) clone. Interestingly, 5 of 6 variants (CrpP1-CrpP5) carried missense mutations in an amino acid position (Gly7) previously defined as essential conferring ciprofloxacin resistance, and decreased ciprofloxacin susceptibility was only associated with the novel CrpP6 protein. In our collection, ciprofloxacin resistance was mainly due to chromosomal mutations in the gyrA and parC genes. However, crpP genes carrying mutations essential for protein function (G7, I26) and associated with a restored ciprofloxacin susceptibility were predominant. Despite the presence of crpP genes is not always associated with ciprofloxacin resistance, the risk of emergence of novel CrpP variants with a higher ability to affect quinolones is increasing. Furthermore, the spread of crpP genes in highly mobilizable genomic islands among related and non-related P. aeruginosa clones alert the dispersion of MDR pathogens in hospital settings.

A vast number of multicriteria decision making methods have been developed to deal with the problem of ranking a set of alternatives evaluated in a multicriteria fashion. Very often, these methods assume that the evaluation among criteria is statistically independent. However, in actual problems, the observed data may comprise dependent criteria, which, among other problems, may result in biased rankings. In order to deal with this issue, we propose a novel approach whose aim is to estimate, from the observed data, a set of independent latent criteria, which can be seen as an alternative representation of the original decision matrix. A central element of our approach is to formulate the decision problem as a blind source separation problem, which allows us to apply independent component analysis techniques to estimate the latent criteria. Moreover, we consider TOPSIS-based approaches to obtain the ranking of alternatives from the latent criteria. Results in both synthetic and actual data attest the relevance of the proposed approach.

In many ranking problems, some particular aspects of the addressed situation should be taken into account in the aggregation process. An example is the presence of correlations between criteria, which may introduce bias in the derived ranking. In these cases, aggregation functions based on a capacity may be used to overcome this inconvenience, such as the Choquet integral or the multilinear model. The adoption of such strategies requires a stage to estimate the parameters of these aggregation operators. This task may be difficult in situations in which we do not have either further information about these parameters or preferences given by the decision maker. Therefore, the aim of this paper is to deal with such situations through an unsupervised approach for capacity identification based on the multilinear model. Our goal is to estimate a capacity that can mitigate the bias introduced by correlations in the decision data and, therefore, to provide a fairer result. The viability of our proposal is attested by numerical experiments with synthetic data

International audience; In several multicriteria decision making problems, it is important to consider interactions among criteria in order to satisfy the preference relations provided by the decision maker. This can be achieved by using aggregation functions based on fuzzy measures, such as the Choquet integral and the multilinear model. Although the Choquet integral has been studied in a large number of works, one does not find the same literature with respect to the multilinear model. In this context, the contribution of this work is twofold. We first provide a formulation of the multilinear model by means of a 2-additive capacity. A second contribution lies in the problem of capacity identification. We consider a supervised approach and apply optimization models with and without regularization terms. Results obtained in numerical experiments with both synthetic and real data attest the performance of the considered approaches.

The Choquet integral has been used as an aggregation operator in the field of multiple criteria decision aiding. Due to its nonlinear nature, the Choquet integral can model interactions between different criteria, such as synergy and redundancy. However, the identification of the Choquet integral parameters is a challenging problem due to its ill-posed nature, which may lead to non-unique solutions. In recent works, this problem has been addressed by considering regularization terms based on sparsity. In this work, this approach is also considered. However, differently from previous studies, in which the Choquet integral is parametrized by means of a fuzzy measure, we propose a novel identification method which exploits sparsity in a transformed domain known as interaction index representation. We provide a set of numerical experiments to assess the proposed method. As a second contribution of the paper, we conduct an identifiability analysis, in which the aim is to search for conditions that ensure that the identification process leads to unique solutions. This analysis is supported by a set of numerical experiments carried out in different scenarios.

International audience; The purpose of this paper is manifold. In a first part, we present a new alternating least squares (ALS)-based method for estimating the matrix factors of a Kronecker product, the so-called Kronecker ALS (KALS) method. Four other methods are also briefly described. In a second part, we consider the design of multiple-input multiple-output (MIMO) wireless communication systems using tensor modelling. Eight systems are presented in a unified way, and their theoretical performance is compared in terms of maximal diversity gain. Exploiting a Kronecker product of symbol and channel matrices, and applying the algorithms introduced in the first part, we propose three semi-blind and two supervised receivers, called Kronecker receivers, for jointly estimating the channel and the transmitted symbols. Necessary identifiability conditions are established. Finally, extensive Monte Carlo simulation results are provided to compare the performance of three tensor-based systems, on the one hand, and of the five proposed Kronecker receivers for the tensor space-time-frequency (TSTF) coding system, on the other hand.

Blind deconvolution is a major theme in signal processing and has been intensely investigated over the last decades. Among its several applications, we can mention the problem of seismic deconvolution and channel equalization in telecommunications. In these two cases, predictive techniques have been studied by different authors, and presented satisfactory results when some suitables conditions were fulfilled. In fact, the predictive deconvolution structure, when associated with the classical mean squared error criterion is only effective when the distortion system is minimum phase. In the case of nonminimum phase systems, it only provides magnitude equalization, but the phase response remains distorted. In order to overcome this problem, we present in this work some interesting results obtained with the use of Lp norms, with p different of 2, as optimization criteria. First we demonstrate that the Lp prediction error filter works as the Maximum Likelihood solution for blind deconvolution when the signal to be recovered has a generalized Gaussian distribution, with i.i.d (identically and independently distributed) samples. From this, we show how the best p can be chosen according to the signal distribution. Then we further investigate the phase response of the Lp filter, emphasizing its potential as well as some limitations in dealing with blind deconvolution, even for nonminimum phase distortion system. Finally, some performance simulations results are provided.

This work proposes the application of independent component analysis to the problem of ranking different alternatives by considering criteria that are not necessarily statistically independent. In this case, the observed data (the criteria values for all alternatives) can be modeled as mixtures of latent variables. Therefore, in the proposed approach, we perform ranking by means of the TOPSIS approach and based on the independent components extracted from the collected decision data. Numerical experiments attest the usefulness of the proposed approach, as they show that working with latent variables leads to better results compared to already existing methods

International audience; In this paper, we address the problem of blind compensation of nonlinear distortions. Our approach relies on the assumption that the input signals are bandlimited. We then make use of the classical result that the output of a nonlinear memoryless system has a wider spectrum than the one of the in- put signal. However, different from previous works, our approach does not assume the knowledge of the input signal's bandwidth. The proposed approached is considered in the development of a two-stage method for blind source separation (BSS) in post- nonlinear (PNL) models. Indeed, once the functions present in the nonlinear stage of a PNL model are compensated, one can apply the well-established linear BSS algorithms to complete the task of separating the sources. Numerical experiments performed in different scenarios attest the viability of the proposal. Moreover, the proposed method is tested in a real situation where the data are acquired by smart chemical sensor arrays.

We derive and analyze two modified Kalman channel estimators (KCE) for time-varying, flat, spatially correlated MIMO channels in systems employing orthogonal space-time block codes: the steady-state KCE, which is less complex than the KCE, and the fading memory KCE, which is more robust to model mismatch.

This paper introduces a novel source coding method for voltage and current signals, called fundamental, harmonic and transient coding method (FHTCM), which is a generalization of the enhanced disturbance compression method (EDCM). The proposed method makes use of notch filtering-warped discrete Fourier transform (NF-WDFT) technique for estimating the parameters (amplitude, frequency, and phase) of the fundamental and harmonic components acquired from power lines so that only the transient components are compressed with wavelet transform (WT) coding technique. For the WT-based compression of transient components, we formulate a minimum description length (MDL) criterion, taking into account the selection of wavelet bases in a dictionary, wavelet decomposition structure, and quantization. Computational simulations have verified that the proposed method outperforms the EDCM as well as the traditional WT-based compression techniques

We highlight challenges associated with using conventional pattern recognition methods to test imagery from a persistent longwave infrared remote sensing experiment, where material surface emissivity and temperature among other factors dominate the apparent radiance observed at the sensor. We also propose a superior data model for the task.

In this paper, we propose a method for blind compensation of a memoryless nonlinear distortion. We assume as prior information that the desired signal admits a sparse representation in a transformed domain that should be known in advance. Then, given that a nonlinear distortion tends to generate signals that are less sparse than the desired one, our proposal is to build a compensating function model that gives rise to a maximally sparse signal. The implementation of this proposal has, as central elements, a criterion built upon an approximation of the $\ell _{0}$ -norm, the use of polynomial functions as compensating structures, and an optimization strategy based on sequential quadratic programming. We provide a theoretic analysis for an $\ell _{0}$ -norm criterion and results considering synthetic data. We also employ the method in an actual application related to chemical analysis via ion-selective electrode arrays.

International audience; This paper is devoted to two classical sparse problems in array processing: Channel estimation and DOA estimation. It is shown after some background and some recent results in l0 optimization how this latter can be used, at the same computational cost, in order to obtain improvement in comparison with l1 optimization for sparse estimation.

This paper introduces a modified version of the discrete multi-tone transceiver (DMT) to increase the data rate in broadband power-line communications (PLC). Basically, a computational-intelligence technique trained by a second-order optimization method is applied to mitigate the high-power impulse noise, while the DMT copes with the severe intersymbol interference observed in power-line channels. The masking of the impulse noise enhances the signal-to-noise ratio in each DMT subchannel. As a result, a high number of bits can be allotted to each subchannel for a given error probability. The simulation results reveal that the proposed PLC-DMT solution outperforms, in terms of the data rate, traditional PLC-DMT over different environments, especially in the presence of additive high-power impulse noise.

International audience; The separation of an underdetermined audio mixture can be performed through sparse component analysis (SCA) that relies however on the strong hypothesis that source signals are sparse in some domain. To overcome this difficulty in the case where the original sources are available before the mixing process, the informed source separation (ISS) embeds in the mixture a watermark, which information can help a further separation. Though powerful, this technique is generally specific to a particular mixing setup and may be compromised by an additional bitrate compression stage. Thus, instead of watermarking, we propose a 'doping' method that makes the time-frequency representation of each source more sparse, while preserving its audio quality. This method is based on an iterative decrease of the distance between the distribution of the signal and a target sparse distribution, under a perceptual constraint. We aim to show that the proposed approach is robust to audio coding and that the use of the sparsified signals improves the source separation, in comparison with the original sources. In this work, the analysis is made only in instantaneous mixtures and focused on voice sources.

This paper introduces a new waveform coding technique, based on wavelet transform, for power quality monitoring purposes. The proposed enhanced data compression method (EDCM) presents a complete adaptive signal processing approach to estimate the fundamental sinusoidal component and separate it from the transient ones in the monitored signal. When these nonstationary components are submitted to the compression technique, the sparse representation property of the wavelet transform provides an improvement in the compression ratio. Also, the degradation inserted by the lossy compression process is minimized. Simulation results confirm the effectiveness of the proposed method when compared to the standard solution, characterized by the compression of the whole monitored signal.

This paper proposes a new structure for split transversal filtering and introduces the optimum split Wiener filter. The approach consists of combining the idea of split filtering with a linearly constrained optimization scheme. Furthermore, a continued split procedure, which leads to a multisplit filter structure, is considered. It is shown that the multisplit transform is not an input whitening transformation. Instead, it increases the diagonalization factor of the input signal correlation matrix without affecting its eigenvalue spread. A power normalized, time-varying step-size least mean square (LMS) algorithm, which exploits the nature of the transformed input correlation matrix, is proposed for updating the adaptive filter coefficients. The multisplit approach is extended to linear-phase adaptive filtering and linear prediction. The optimum symmetric and antisymmetric linear-phase Wiener filters are presented. Simulation results enable us to evaluate the performance of the multisplit LMS algorithm.

This work proposes a framework to determine the optimal Wiener equalizer by using an artificial immune network model together with the constant modulus (CM) cost function. This study was primarily motivated by recent theoretical results concerning the CM criterion and its relation to the Wiener approach. The proposed immune-based technique was tested under different channel models and filter orders, and benchmarked against a procedure using a genetic algorithm with niching. The results demonstrated that the proposed strategy has a clear superiority when compared with the more traditional technique. The proposed algorithm presents interesting features from the perspective of multimodal search, being capable of determining the optimal Wiener equalizer in most runs for all tested channels.

This report summarizes our observations using the longwave infrared (LWIR) hyperspectral data subset of the Spectral and Polarimetric Imagery Collection Experiment (SPICE) dataset, which emphasizes the inherent challenges associated with applying autonomous spectral pattern recognition methods to LWIR hyperspectral imagery, as well as data quality and qualitative validation of expected atmospheric spectral features. SPICE is a collaborative effort between the US Army Research Laboratory, US Army Armament Research, Development, and Engineering Center, and the US Air Force Institute of Technology. It focused on the long-term LWIR hyperspectral and polarimetric data collection of a target site, under various weather and non-ideal conditions, and exploitation of this dataset for pattern recognition tasks. We concluded from this effort that the quality of SPICE hyperspectral LWIR data is categorically comparable to other datasets recorded by a different sensor of similar specifications and is adequate for algorithmic research. The vast dataset and associated ground-truth information are a great asset for research.

International audience; As shown recently, the interference problem typical of ion-selective electrodes can be dealt with via smart arrays adjusted by blind source separation methods. In this letter, we resume this study and show that such an approach can be applied even when faced with a limited number of samples acquired through flow-injection analysis.

Recent work has shown that, besides providing a conventional stacked section in which reflections are enhanced, the Common-Reflection-Surface (CRS) stacking method may be used to obtain a stacked section mainly composed of diffraction events. Therefore, CRS can be applied to separate reflection and diffraction events in the resulting zero-offset (ZO) sections. However, in practice, there is still mutual interference in the sections obtained by the stacking process. To improve the quality of separation, we propose an additional step based on a signal processing framework known as blind source separation (BSS). The adopted BSS method relies on the concept of sparsity, which suits well the overall characteristic of the reflection events. Numerical experiments with synthetic and field data show that the proposed additional step leads to better results with respect to the stacked section generated by CRS.

An extension of the field of fast least-squares techniques is presented. It is shown that the adaptation gain, which is updated with a number of operations proportional to the number of transversal filter coefficients, can be used to update the coefficients of a linearly constrained adaptive filter. An algorithm that is robust to round-off errors is derived. It is general and flexible. It can handle multiple constraints and multichannel signals. Its performance is illustrated by simulations and compared with the classical LMS-based Frost (1972) algorithm.

Space-time codes are an effective and practical way to exploit the benefits of spatial diversity in multiple-input, multiple-output (MIMO) systems. As the name suggests, the coding of space-time codes is performed in both the spatial and the temporal domains, introducing correlation between signals transmitted by different antennas at different time instants. At the receiver, this space-time correlation is exploited to mitigate the detrimental effects of fading and to improve the quality of received signal. The use of space-time codes allows the system to profit from transmit diversity without any power increment and with no channel knowledge at the transmitter. Among the existing space-time coding schemes, orthogonal space-time block codes (OSTBCs) (Alamouti, 1998; Duman & Ghrayeb, 2007; Larsson & Stoica, 2003; Tarokh et al., 1999) are of particular interest because they achieve full diversity at low receiver complexity. More specifically, the maximum likelihood (ML) receiver for OSTBCs consists of a linear receiver followed by a symbol-by-symbol decoder. In order to correctly decode the received signals, the ML receiver for OSTBCs must have perfect channel knowledge. Unfortunately, this channel information is not normally available to the receivers; therefore channel estimation techniques are essential for the system to work properly. When the channel is static, methods such as those presented in (Vucetic & Yuan, 2003) and (Larsson et al., 2003) can be successfully used. However, the channel may be time-varying due to the mobility of the transmitter and/or receiver, to changes in the environment, or to carrier frequency mismatch between transmitter and receiver. In these cases, the estimation algorithm must be able to track the channel variations. One of the most widely known approaches to channel tracking is Kalman filtering (Enescu et al., 2007; Kailath et al., 2000; Komninakis et al., 2002; Piechocki et al., 2003; Simon, 2006). An important characteristic of the Kalman filter (KF) is its inherent ability to deal with nonstationary environments. In (Komninakis et al., 2002) for instance, a Kalman filter that uses the outputs of a minimum mean squared error (MMSE) decision-feedback equalizer (DFE) is developed to track Rice MIMO frequency-selective channels. Channel estimation using Kalman filters for MIMO-OFDM systems is studied in (Enescu et al., 2007; Piechocki et al., 2003). The use of Kalman filters to estimate channels in space-time block coded MIMO systems is also developed in the literature. In (Liu et al., 2002), a KF is used to estimate fast flat fading MIMO channels in Alamouti-based schemes. Therefore, it is limited to the case of two transmit 13

ISBN 978-3-642-15994-7, Softcover; International audience; This work deals with the problem of source separation in overdetermined linear-quadratic (LQ) models. Although the mixing model in this situation can be inverted by linear structures, we show that some simple independent component analysis (ICA) strategies that are often employed in the linear case cannot be used with the studied model. Motivated by this fact, we consider the more complex yet more robust ICA framework based on the minimization of the mutual information. Special attention is given to the development of a solution that be as robust as possible to suboptimal convergences. This is achieved by defining a method composed of a global optimization step followed by a local search procedure. Simulations confirm the effectiveness of the proposal.

In this paper we propose low-complexity algorithms for estimating flat, time-varying and spatially correlated MIMO channels. The proposed estimators employ Kalman filters to track the channel in orthogonal space-time block coded systems. After developing a state-space model for spatially correlated MIMO channels, we show that the proposed estimators can be simplified by using the orthogonality inherent to orthogonal space-time block codes. We also show that the channel estimates provided by the proposed algorithms correspond to weighted sums of instantaneous maximum likelihood channel estimates. For constant modulus signal constellations, we reduce the receiver complexity even more by proposing a steady-state Kalman filter. Simulation results indicate that there is no significant performance degradation of the steady-state filter compared to the first proposed algorithm and that both algorithms outperform conventional adaptive filters.

International audience; Usually, source separation in Post-Nonlinear (PNL) models is achieved via one-stage methods, i.e. the two parts (linear and nonlinear) of a PNL model are dealt with at the same time. However, recent works have shown that the development of two-stage techniques may simplify the problem. Indeed, if the nonlinear stage can be compensated separately, then, in a second moment, one can make use of the well-established source separation algorithms for the linear case. Motivated by that, we propose in this work a novel two-stage PNL method relying on the assumption that the sources are bandlimited signals. In the development of our method, special care is taken in order to make it as robust as possible to noise. Simulation results attest the effectiveness of the proposal.

ao pode levar a jogos com m ´ ultiplos pontos de equilde Nash ou mesmo a jogos onde n˜ ao h´ a nenhum ponto de equilSimulac ¸ ˜ oes em cen´ arios DSL representativos s˜ ao apresentadas de modo a comparar o algoritmo com restric ¸ ˜ ao discreta de alocac ¸ ˜ ao de bits com o bem conhecido algoritmo iterative water-filling. Palavras-Chave— Digital subscriber lines, Gerenciamento espectral distribuSistemas multiportadora. Abstract— This work analyzes the impact of the discrete bitloading constraint on the convergence of a iterative algorithm for distributed spectrum management in a DSL environment. The power control problem is formulated as a noncooperative game in which the users unilaterally optimize their transmit power and bit allocations in order to maximize the rate. In this context, it is shown that the inclusion of the discrete bitloading constraint in the optimization problem may lead to game with multiple Nash equilibria or even with no equilibrium, depending on the characteristics of the communication environment. Simulations in representative DSL scenarios are presented and the results of the iterative discrete bitloading are compared with those of the well known iterative water-filling algorithm.

We propose a method for source separation of convolutive mixture based on nonlinear prediction-error filters. This approach converts the original problem into an instantaneous mixture problem, which can be solved by any of the several existing methods in the literature. We employ fuzzy filters to implement the prediction-error filter, and the efficacy of the proposed method is illustrated by some examples.

In this work, we propose an evolutionary-like approach to the problem of blind adaptive spatial filtering that is based on the decision-directed criterion and on the dopt-aiNet, an artificial immune network conceived to perform multimodal search in dynamic environments. The proposal was tested under static and time-varying undermodeled channel models, and, in all cases, its ability to find and track a solution close to the Wiener global optimum was attested. The obtained results reveal that the dopt-aiNet may decisively enhance the performance of adaptive arrays in scenarios built from elements that are representative of some aspects of real-world communication systems.

In this paper, we make use of a blind adaptive linear predictor for channel shortening in single input multiple output (SIMO) channels. We compare our approach to the so-called MERRY blind channel shortener. We assess through simulations that our proposed approach provides faster convergence rate and it better exploits the spatio-temporal diversity present in the SIMO channels.

Neste artigo são abordados critérios usados para resolver o problema da equalização cega também conhecida como autodidata. Consideram-se os critérios clássicos do módulo constante e o do Shalvi-Weinstein. Apresentaremos os principais resultados existentes na literatura e alguns resultados mais recentes, que dizem respeito ao estudo do algoritmo do módulo constante generalizado (GCMA) e à comparação entre os critérios citados e destes com o critério de Wiener. In this work, we study the criteria used to solve the blind equalization problem. Two approaches are considered in detail: the constant modulus and the Shalvi-Weinstein criteria. In the course of our exposition, a more recent and less studied technique, the generalized constant modulus criterion, is also discussed. Some of the most important results found in the literature are presented together with some recent contributions related to the comparison between blind criteria and between unsupervised techniques and the Wiener criterion.

Due to its universal approximation capability, the multilayer perceptron (MLP) neural network has been applied to several function approximation and classification tasks. Despite its success in solving these problems, its training, when performed by a gradient-based method, is sometimes hindered by the existence of unsatisfactory solutions (local minima). In order to overcome this difficulty, this paper proposes a novel approach to the training of a MLP based on a simple artificial immune network model. The application domain for assessing the performance of the proposed technique is that of digital communications, in particular, the problems of channel equalization and pre-distortion. The obtained simulation results demonstrate that the proposal is capable of efficiently solving the problems tackled

In this paper we discuss the space-time channel modeling aiming applications of adaptive antennas in mobile communication systems. Some classical approaches for this sake are reviewed and the underlying assumptions discussed. A model that unifies these approaches by including them as particular cases is then presented. Simulation results on adaptive antenna arrays in a TDMA cellular system illustrate the impact of the adoption of each modeling approach in its performance as well as the relationship among them.

This work has a twofold aim: to present a numerical analysis of the Hodgkin–Huxley model in a nonsmooth excitation scenario — which is both challenging and theoretically relevant — and to use the established framework as a basis for testing a method to search for specific oscillating patterns in dynamical systems. The analysis is founded on classical qualitative methods — bifurcation diagrams, phase space and spectral analysis — and on the calculation of the system Lyapunov spectrum. This calculation is carried out by means of an algorithm particularly suited to deal with nonsmooth excitation and the complexity of the state equations. The obtained Lyapunov exponents are then used to build a robust cost function (invariant with respect to the initial conditions or specific trajectories in a given basin of attraction) for seeking predefined dynamical patterns that are optimized using the particle swarm optimization algorithm. This bioinspired method possesses two desirable features: it has a significant global search potential and does not demand cost function manipulation. The proposed approach, which was tested here in different representative scenarios for the Hodgkin–Huxley model, has a promising application potential in general dynamical contexts and can also be a valuable tool in the planning of drug administration and electrical stimulation of neuronal and cardiac cells.

Low-complexity blind turbo receivers composed of a soft-feedback equaliser (SFE) and fast least-squares (FLS) channel estimators are proposed. To reduce the complexity of the SFE computation, a function approximation is proposed instead of numerical algorithms in a specific step of the equaliser's computation. Concerning FLS channel estimators, a specific filter parameters initialisation procedure is proposed in each turbo iteration to avoid possible numerical instabilities. Estimation of noise variance is also considered. The proposed scheme can perform as the turbo equaliser with perfect channel knowledge from a certain signal-to-noise ratio.

The paper introduces a new objective method for speech quality assessment called Fourier-lapped multilayer perceptron (FLMLP). This method uses an overcomplete transform based on the discrete Fourier transform (DFT) and modulated lapped transform (MLT). This transform generates the DFT and the MLT speech spectral domains from which several relevant perceptual parameters are extracted. The proposed method also employs a multilayer perceptron neural network trained by a modified version of the scaled conjugated gradient method. This neural network maps the perceptual parameters into a subjective score. The numerical results show that FLMLP is an effective alternative to previous methods. As a result, it is worth stating that the techniques here described may be potentially useful to other researches facing the same kind of problem.

Orientador: Helio Waldman Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia de Campinas Resumo: Não informado. Abstract: Not informed. Mestrado Mestre em Engenharia Elétrica