Orientador: João Marcos Travassos Romano / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-04T07:52:07Z (GMT). No. of bitstreams: 1 Attux_RomisRibeirodeFaissol_D.pdf: 4889703 bytes, checksum: 022efe9d9ffb4a1629612763bc2c2b3d (MD5) Previous issue date: 2005 / Resumo: O objetivo deste trabalho é investigar a aplicação de estruturas não-lineares e de técnicas de otimização baseadas em computação evolutiva a problemas de equalização e identificação de canal. O relato se divide em duas partes: a primeira voltada à análise dos fundamentos do problema de filtragem, e a segunda, à apresentação de novas abordagens para sua solução. A primeira parte, inaugurada pelas noções primordiais de comunicação, abrange os diferentes aspectos do projeto de um filtro. Permeia toda a exposição uma idéia fundamental: o estabelecimento de um paradigma genérico de filtragem adaptativa. Na segunda parte, apresentamos contribuições originais que se encaixam de diversas formas no espírito desse paradigma. Os problemas abordados são: equalização linear cega, equalização e pré-distorção baseadas em redes neurais, identificação cega, identificação de plantas recursivas, busca cega do receptor de máxima verossimilhança e equalização não-linear cega baseada em predição. Tais propostas, além de possuírem um valor intrínseco, podem ser entendidas como um corpus de evidências da validade das idéias unificadoras pertencentes ao arcabouço teórico erigido / Abstract: The objective of this work is to investigate the use of nonlinear structures and optimization techniques based on evolutionary computation in channel equalization and identification problems. The text is structured according to a twofold division: the first part is dedicated to the analysis of the foundations of the filtering problem, and the second, to the exposition of new approaches to its solution. The first part, whose starting point is the primordial idea of communication, encompasses the various aspects of the design of a signal-processing device. The entire explanation is permeated by a fundamental idea: the constitution of a generic paradigm for adaptive filtering. In the second part, a number of original contributions are presented that are related in various ways to the essence of this paradigm. These contributions involve the problems of blind linear equalization, neural network-based equalization and pre-distortion, blind and supervised identification of feedforward and feedback plants, maximum-likelihood sequence estimation and nonlinear blind equalization based on prediction. Such proposals, apart from possessing an intrinsic value, can also be understood as a corpus of evidences of the validity of the unifying ideas belonging to the established theoretical framework / Doutorado / Telecomunicações e Telemática / Doutor em Engenharia Elétrica

Digital signal processing (DSP) is widely used in many applications spanning the spectrum from audio processing to image and video processing to radar and sonar processing. At the core of digital signal processing applications is the digital filter which are implemented in two ways, using either finite impulse response (FIR) filters or infinite impulse response (IIR) filters. The primary difference between FIR and IIR is that for FIR filters, the output is dependent only on the inputs, while for IIR filters the output is dependent on the inputs and the previous outputs. FIR filters also do not sur from stability issues stemming from the feedback of the output to the input that aect IIR filters. In this thesis, an architecture for FIR filtering based on distributed arithmetic is presented. The proposed architecture has the ability to implement large FIR filters using minimal hardware and at the same time is able to complete the FIR filtering operation in minimal amount of time and delay when compared to typical FIR filter implementations. The proposed architecture is then used to implement the fast affine projection adaptive algorithm, an algorithm that is typically used with large filter sizes. The fast affine projection algorithm has a high computational burden that limits the throughput, which in turn restricts the number of applications. However, using the proposed FIR filtering architecture, the limitations on throughput are removed. The implementation of the fast affine projection adaptive algorithm using distributed arithmetic is unique to this thesis. The constructed adaptive filter shares all the benefits of the proposed FIR filter: low hardware requirements, high speed, and minimal delay.

In this thesis we describe a method of mapping one-dimensional and multidimensional filter algorithms onto systolic architectures using the z-domain approach. In this approach the filter algorithm is first transformed into its corresponding z-domain equivalent and recursive expressions similar to single assignment codes are derived using Horner's rule or other polynomial evaluation techniques. By obtaining different recursive expressions, different systolic structures can be derived. The characteristics of these structures can easily be deduced from the recursive expressions. The multidimensional filters derived are modular and hierarchical, i.e., the three-dimensional structures are obtained from the two-dimensional ones which are in turn obtained from one-dimensional structures. In considering the design of any array processor, it is important to consider the design of the processing elements involved. The most important and demanding operation in these elements is the multiplication. Four different multipliers are designed in which the number of operations required to produce the desired result is reduced. The reduced number of operations along with the advantages of very-large-scale integration technology in terms of increased device density and faster switching make these multipliers potential candidates in high-speed signal processing applications. The first multiplier is an area-efficient multiplier that uses approximately 50% of the area of a full parallel multiplier. In this multiplier only the units yielding the most significant part of the product are used. In addition, a correction unit is incorporated to minimize the error resulting from circumventing the use of units yielding the least significant part of the product. The second multiplier is based on the modified octal Booth algorithm in which four-bit segments of the multiplier are scanned and corresponding operations effected on the multiplicand. The third multiplier is a diminished-1 multiplier that finds application in the Fermat number-theoretic transform. In this multiplier the use of a translator is circumvented and a novel technique for translation is incorporated in the multiplier structure. The fourth multiplier is one that performs an inner-product operation without the use of an accumulator thereby resulting in increased speed and reduced area. Finally we discuss the VLSI implementations of three of the multipliers mentioned above, a second-order digital filter, and a single processing element that can be used as a basic unit in designing one-dimensional and multidimensional digital filters. Some associated problems in digital-filter structure. viz., the quantization and overflow limit-cycle oscillations; have been taken into consideration and ways have been suggested for their elimination. / Graduate

Orientador: Amauri Lopes / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-09T06:28:20Z (GMT). No. of bitstreams: 1 Krummenauer_Rafael_M.pdf: 1711180 bytes, checksum: e9aa73aac9d705c24c2bec02b202f76e (MD5) Previous issue date: 2007 / Resumo: Esta dissertação trata do problema de estimação de direção de chegada (DOA) de ondas planas usando um arranjo linear uniforme de sensores. Estamos interessados em situações nas quais a relação sinal-ruido 'e baixa e o espaçamento angular entre as fontes de sinal 'e pequeno. Baseamos nossa proposta nos m'etodos MODE, MODEX e MODEX Modificado, que sao metodos eficientes existentes na literatura. Inspirados em conceitos de filtragem linear e no criterio da maxima verossimilhança, propomos um procedimento que ameniza o efeito do ruido no resultado da estimação. Este procedimento consiste em filtrar os dados recebidos e modificar adequadamente a função de verossimilhan¸ca utilizada no processo de obtenção das estimativas. Simulações numericas mostram que o desempenho do metodo proposto 'e melhor que aqueles correspondentes aos m'etodos MODE, MODEX e MODEX Modificado, alcançando menores valores de erro quadratico medio e de polarização / Abstract: This work deals with the problem of estimating the direction of arrival (DOA) of plane waves using a uniform linear array of sensors. We are concerned with situations where the signal-to-noise ratio is low and the signal sources are spatially close. Our proposal is based on MODE, MODEX and Modified MODEX, that are efficient methods proposed in the literature. Inspired in concepts of linear filtering and in the maximum likelihood criterion, we propose a procedure that reduces the effect of noise in the estimation result. This procedure consists on filtering the received data and on modifying the likelihood function used to obtain the estimates. Numerical simulations show that the performance of the proposed method is better than those of MODE, MODEX and Modified MODEX methods, achieving lower mean square error and lower bias / Mestrado / Telecomunicações e Telemática / Mestre em Engenharia Elétrica

Orientador: João Bosco Ribeiro do Val / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação / Made available in DSpace on 2018-08-25T04:12:10Z (GMT). No. of bitstreams: 1 Frencl_VictorBaptista_D.pdf: 1574127 bytes, checksum: 86e39d74bf9c9e7734aa764b39aaac1a (MD5) Previous issue date: 2014 / Resumo: O problema de rastreamento de alvos é tratado de diversas formas na literatura, seja elaborando modelos matemáticos mais eficientes na reprodução da dinâmica de movimentos, seja na construção de filtros estocásticos que realizem estimativa de estados, como de posição e velocidade. Quando se trata do rastreamento em que os alvos de interesse são diversos indivíduos em movimento, a literatura não possui estudos específicos. Dessa forma, o objetivo principal da tese é aprofundar o conhecimento de rastreamento de indivíduos. Neste cenário, existe um número elevado e variável de alvos, que podem surgir de forma espontânea, agrupar-se ou separar-se, além de alarmes falsos imersos nas medidas. Estudou-se a teoria dos Conjuntos Aleatórios Finitos, cujo tratamento matemático se dá através do chamado Cálculo Multi-Alvo. Os filtros estocásticos também foram estudados sobre este ponto de vista, sendo os filtros PHD e GM-PHD os principais. Criada essa base teórica, três propostas baseadas nesse problema foram apresentadas: Modelos de Movimentação de Indivíduos, Simulador de Trajetórias de Indivíduos e Modelos Dinâmicos para Filtragem Estocástica. A primeira das propostas consiste em construir perfis probabilísticos de movimentação para cada um dos indivíduos. A segunda envolve a criação de um simulador de trajetórias de indivíduos que seja o mais verossímil possível em relação às trajetórias reais de uma pessoa, em cenários com variações de terreno, classificados pela dificuldade de locomoção. E finalmente, a terceira proposta tem como objetivo criar um modelo dinâmico combinado e modificado em relação a modelos encontrados na literatura para ser inserido no processo de filtragem estocástica. Ao final, alguns testes e simulações foram realizados, de tal forma a testar o desempenho de filtros e analisar o comportamento dos modelos matemáticos e dos perfis probabilísticos propostos / Abstract: The problem of target tracking is handled in different ways in the literature, either developing more efficient mathematical models to reproduce the dynamics of movements, or building stochastic filters that perform state estimation, such as position and velocity. When it comes to target tracking where the targets of interest are many individuals in motion, the literature lacks on specific studies. Thus, the main objective of the thesis is to deepen the knowledge of individuals tracking. In this scenario, there is a large and variable number of targets, which may arise spontaneously, group together or separate, in addition to measures immersed in false alarms. A study of the Random Finite Sets theory was made, whose mathematical treatment is through the so-called Multi-Target Calculus. Stochastic filters were also studied on this point of view, where the PHD and the GM-PHD filters are the main ones. After created the theoretical basis, three proposals based on this problem were presented: Motion Models for Individuals, a Simulator for Individuals Trajectories and Dynamic Models for Stochastic Filtering. The first proposal is based on building a motion probabilistic shape for each individual. The second proposal involves the creation of a trajectory simulator for individuals to be as plausible as possible to the real movements of a person, in scenarios with variations of terrain, ranked by locomotion difficulty. And finally, the third proposal aims to create a combined and modified dynamic model from models found in the literature, to be inserted in the stochastic filters. Finally, several tests and simulations were made in such a way to test the filters performances and analyze the behavior of the proposed mathematical models and the motion probabilistic shapes / Doutorado / Automação / Doutor em Engenharia Elétrica

Thesis (M.Sc. (Mathematics)) -- University of Limpopo, 2021 / We study balanced filters and balanced z-filters considered by Carlson in [20] and [21] in topological spaces. We consider closed filters which are open-generated and open filters which are closed-generated. We show that a closed filter is open-generated precisely if it is a minimal balanced closed filter and that an open filter is closed-generated precisely when it is a minimal balanced open filter. For a completely regular topological space X, we study balanced z-filters and show that there is a one-to-one correspondence between the nonempty closed sets of βX and the balanced z-filter on X. By dualising closed filters we obtain ideals which then enables us to put some of the results in the context of frames. Dube in [28] has shown that a frame is normal if and only if its closed-generated filters are precisely the stably closed-generated ones. By dualisation we show that a frame is extremally disconnected if and only if its open-generated ideals are precisely the stably open-generated ones. We show that there is one-to-one correspondence between points of βL and the balanced ideals of Coz L. Furthermore we study nearness frames and show that the locally finite nearness frames strictly contain the Pervin nearness frames and the two coincide if the locally finite nearness frames are totally bounded. For perfect extension h : M → L of L, we show that a point p of M is a remote point if and only if Ip = {a ∈ L | h∗(a) ≤ p}. / University of Limpopo and DST - NRF Centre of Excellence in Mathematical and Statistical Sciences (CoE - MaSS)

Orientador: Amauri Lopes / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação / Made available in DSpace on 2018-08-14T21:35:53Z (GMT). No. of bitstreams: 1 Gomes_MarcoAurelioCazarotto_M.pdf: 1012758 bytes, checksum: 0c2ca6c09e4123b277735dca0f50a107 (MD5) Previous issue date: 2009 / Resumo: Abordamos o problema de estimação de direção de chegada (DOA) de ondas planas usando um arranjo de sensores. Na literatura encontramos diversos estimadores para DOA, porém estamos considerando apenas os estimadores de Máxima Verossimilhança (ML) que geram candidatas à estimativa DOA e selecionam as melhores através do critério ML. Também estamos interessados em situações em que o espaçamento angular entre as fontes de sinal é pequeno e a relação sinal-ruído é baixa. Nesse caso temos uma degradação de desempenho associada ao efeito de limiar. Mostramos que este problema pode ser amenizado reduzindo o ruído presente na matriz de covariância dos dados recebidos (snapshots) utilizada para a seleção das candidatas. Propomos então modificar o processo de seleção de candidatas, utilizando uma nova matriz de covariância dos snapshots, calculada após uma filtragem ótima dos dados através de um filtro FIR multibanda. Propomos também modificar a função custo ML para adequá-la às dimensões da matriz de covariância filtrada e para isso apresentamos 3 opções de modificação. As simulações mostram que nossa proposta tem melhor desempenho que os métodos conhecidos, reduzindo significativamente a relação sinal-ruído de limiar. / Abstract: We approached the estimation of direction of arrival (DOA) of plane waves using an array of sensors. In the literature there are several DOA estimators, but we considered only the maximum likelihood (ML) estimators that generate candidates for DOA estimation and select the best one through an ML criterion. We also considered situations where the signal sources are spatially closely spaced and the signal-to-noise ratio is low. In these cases a performance degradation associated with the threshold effect occur. We demonstrated that we can improve the estimation performance by reducing the noise in the received data covariance matrix used to select the candidates. Then we proposed to modify the selection process using a new data covariance matrix, computed after an optimum multiband FIR filtering of the received data. We also proposed to modify the ML cost function to adapt it to the dimensions of the new covariance matrix and we considered 3 alternatives of modification. Some simulations showed that our proposal has better performance than known DOA methods, significantly reducing the threshold SNR. / Mestrado / Telecomunicações e Telemática / Mestre em Engenharia Elétrica

Esta dissertação trata sobre o projeto, implementação e avaliação de um filtro híbrido para supressão de ruído em sistemas de alta potência. Seu desempenho será otimizado para reduzir a magnitude de ruídos impulsivos, que são comuns em dispositivos de alta potência, como inversores de frequência que controlam motores trifásicos. O filtro híbrido proposto é avaliado empiricamente em um inversor de frequência que é controlado por um sistema embarcado. A abordagem proposta é comparada com abordagens clássicas de filtragem digital como média móvel, filtro de resposta finita ao impulso (FIR) e filtro de resposta infinita ao impulso (IIR). / This work presents the project, implementation and evaluation of a hybrid filter used for noise supressing in high power switching converters. It is optimized to reduce impulsive noise that is commonly present in high power devices like frequency inverters that control three-phase motors. The hybrid filter is evaluated empirically in a frequency inverter that is controlled by an embedded system. This approach is compared to classical ones, like the moving average, the finite impulse response (FIR) and the infinite impulse response (IIR) filters.

Esta dissertação trata sobre o projeto, implementação e avaliação de um filtro híbrido para supressão de ruído em sistemas de alta potência. Seu desempenho será otimizado para reduzir a magnitude de ruídos impulsivos, que são comuns em dispositivos de alta potência, como inversores de frequência que controlam motores trifásicos. O filtro híbrido proposto é avaliado empiricamente em um inversor de frequência que é controlado por um sistema embarcado. A abordagem proposta é comparada com abordagens clássicas de filtragem digital como média móvel, filtro de resposta finita ao impulso (FIR) e filtro de resposta infinita ao impulso (IIR). / This work presents the project, implementation and evaluation of a hybrid filter used for noise supressing in high power switching converters. It is optimized to reduce impulsive noise that is commonly present in high power devices like frequency inverters that control three-phase motors. The hybrid filter is evaluated empirically in a frequency inverter that is controlled by an embedded system. This approach is compared to classical ones, like the moving average, the finite impulse response (FIR) and the infinite impulse response (IIR) filters.