No description available.

Wavelets have received considerable attention in recent general signal processing, image processing, and pattern recognition literature, as a new method of signal analysis. This marks a transition in wavelet study from theoretical investigation to application-driven research. In this paper, wavelets and wavelet transformations are presented in a context intended to be appropriate as a first exposure to the engineer. The wavelet transform, more specifically the discrete wavelet transform, and its relationship to multiresolution analysis is then explored in a framework familiar to those versed in multirate digital signal processing concepts. Elements of the perspective offered by wavelet analysis, in contrast to the features of more conventional Fourier techniques, are examined. General procedures for wavelet-based signal processing applications are discussed and the specific application of reducing noise in audio signals examined. Within the context of this application, considerations unique to wavelet analysis are revealed and trade-offs analyzed. Finally, the results obtained from implementing the noise reduction system are presented and extensions to the technique proposed. / Master of Science

Blind  people use echoes to detect  objects  and  to  find their  way, the  ability  being known as human echolocation.   Previous  research  have found some of the  favorable  conditions  for the  detection  of the object,  with  many  factors  yet  to  be analyzed  and  quantified.    Studies  have  also shown  that blind people are more efficient than  the  sighted  in echolocating,  with  the  performance  varying  among  the individuals.   This  motivated the  research  in human  echolocation  to move in a new direction  to get a fuller understanding for the high detection  of the blind.  The psychoacoustic  experiments solely cannot determine  how the superior echo detection  of the blind listeners should be attributed to perceptual or physiological causes.  Along with the perceptual results it is vital to know how the sounds are processed in the  auditory system.   Hearing  research  has led to the  development of several auditory  models by combining  the  physiological  and  psychological  results  with  signal  analysis  methods.    These  models try  to describe how the auditory system  processes the signals.  Hence, to analyze how the sounds are processed for the high detection  of the blind, auditory  models available  in the literature were used in this thesis.  The results  suggest  that repetition pitch  is useful at  shorter  distances  and is determined from the peaks in the temporal  profile of the autocorrelation function computed  on the neural activity pattern. Loudness attribute also plays a role in providing information  for the listeners to echolocate at shorter  distances.  At longer distances  timbre  aspects such as sharpness  information  might be used by the listeners  to detect  the objects.  It was also found that the repetition pitch,  loudness and sharpness attributes in turn  depend on the room acoustics  and type of the stimuli  used.  These results  show the fruitfulness  of combining  results  from different  disciplines  through  a mathematical framework  given by signal analysis.

Students at Uppsala University have for some years been given the opportunity to take courses in subjects directly, or indirectly, related to the fields of signal processing and signal analysis. According to the directors of these courses, a considerable number of students are recurringly having difficulties grasping different concepts related to this field of study. This report covers a tool that easily allows teachers to visualize and listen to different manipulations of signals, which should help students get an intuitive understanding of the subject. Features of the system include multiple kinds of analog filters, sampling with variable settings and zero-order hold reconstruction. The finished system is flexible, tunable and modifiable to the teachers every need, making it usable for a wide variety of courses involving signal processing. The system meets its requirements even though individual components’ results de- viate slightly from ideal values. / Studenter vid Uppsala Universitet har, under ett antal år, givits möjligheten att läsa kurser inom ämnen direkt, eller indirekt, relaterade till signalbehandling/signalanalys. Enligt kursansvariga för dessa kurser har en ansenlig andel av studenterna svårigheter med att förstå en del av de begrepp och fenomen som förekommer under kurserna. Denna rapport behandlar ett verktyg som ger lärare i dessa kurser möjlighet att på ett enkelt sätt visualisera och lyssna på olika manipulationer av signaler, vilket bör hjälpa studenterna bygga en intuition för ämnet. Systemets olika funktioner inkluderar flera olika typer av analoga filter, sampling med olika inställningar, och så kallad ’Zero-Order-Hold’ rekonstruktion. Det resulterande systemet är flexibelt, inställbart och modifierbart till användarens behov, vilket gör det applicerbart i flera kurser som innefattar signalbehandling/analys. Systemet möter kraven som ställs, även fast resultaten hos individuella komponenter avviker aningen från ideala värden.

The objective of this thesis was to evaluate radio receivers that had been previously acquired for other tasks by the Swedish Defence Materiel Administration, for additional service in a radio direction finding (DF) system. The antenna, calibration signal generator and DF algorithm were provided. As the receivers used the VITA 49 frame format, much of the work involved the interpretation of the frame structure and contents. The other main challenges were to put the receivers into the correct internal state, to compensate the analogue phase errors due to cabling, and to correlate the digital frames according to the time stamps. The work was performed partly as a literature study and, during implementation, by using an investigative trial and-error method. Collection of reference signals, for later offline analysis, was made at the Swedish Defence Research Agency. The findings of the evaluation was that the receivers were phase coherent and that the VITA 49 frame format was suitable for DF use. An apparently systematic error of unknown origin rendered the DF results inaccurate, even though the precision of the results was high. The automatic gain control (AGC) of the receivers did not perform according to norm. If the systematic error is compensated for, and the AGC problem is eliminated, the receivers will function as a DF system using the manufacturer's instructions, with the provided antenna. The performance of the receivers in this setting was not a primary concern for the thesis, but was deemed to be acceptable. Suggestions for further development include finding the source of the systematic error, an automatic calibration function, a more thorough performance analysis, and code optimizations using multiple threads.

Although the primary purpose of the signal received by amobile handset or smartphone is to enable wireless communication, the information extracted can be reused to provide a number of additional services. Two such services discussed in this thesis are: mobile speed estimation and signal anomaly detection. The proposed algorithms exploit the propagation environment specific information that is already imprinted on the received signal and therefore do not incur any additional signalling overhead. Speed estimation is useful for providing navigation and location based services in areas where global navigation satellite systems (GNSS) based devices are unusable while the proposed anomaly detection algorithms can be used to locate signal faults and aid spectrum sensing in cognitive radio systems. The speed estimation algorithms described within this thesis require a receiver with at least two antenna elements and a wideband radio frequency (RF) signal source. The channel transfer function observed at the antenna elements are compared to yield an estimate of the device speed. The basic algorithm is a one-dimensional and unidirectional two-antenna solution. The speed of the mobile receiver is estimated from a knowledge of the fixed inter-antenna distance and the time it takes for the trailing antenna to sense similar channel conditions previously observed at the leading antenna. A by-product of the algorithm is an environment specific spatial correlation function which may be combined with theoretical models of spatial correlation to extend and improve the accuracy of the algorithm. Results obtained via computer simulations are provided. The anomaly detection algorithms proposed in this thesis highlight unusual signal features while ignoring events that are nominal. When the test signal possesses a periodic frame structure, Kullback-Leibler divergence (KLD) analysis is employed to statistically compare successive signal frames. A method of automatically extracting the required frame period information from the signal is also provided. When the signal under test lacks a periodic frame structure, information content analysis of signal events can be used instead. Clean training data is required by this algorithm to initialise the reference event probabilities. In addition to the results obtained from extensive computer simulations, an architecture for field-programmable gate array (FPGA) based hardware implementations of the KLD based algorithm is provided. Results showing the performance of the algorithms against real test signals captured over the air are also presented. Both sets of algorithms are simple, effective and have low computational complexity – implying that real-time implementations on platforms with limited processing power and energy are feasible. This is an important quality since location based services are expected to be an integral part of next generation cognitive radio handsets.

A análise de sinais mioelétricos provenientes do tubo gastro-intestinal de animais de laboratório (ratos), conseguidos por meio de eletrodos cronicamente implantados, é peça-chave no entendimento das desordens associadas ao sistema digestivo. Esta análise enfrenta consideráveis dificuldades quando realizadas por métodos clássicos, em especial os baseados na transformada de Fourier. A interação de várias componentes mioelétricas torna muito complicado e trabalhoso o acompanhamento destes sinais ao longo do tubo digestivo e a obtenção de parâmetros típicos como a velocidade de propagação entre eletrodos. Estuda-se aqui uma alternativa mais nova e promissora: a transformada Wavelet. Utilizando esta ferramenta matemática, torna-se possível obter uma melhor resolução tempo-freqüencial dos sinais estudados, permitindo encontrar padrões referentes à propagação do sinal mesmo em leituras ruidosas e compostas de várias freqüências. Foram analisados 82 leituras de 9 animais normais do Laboratório de Investigação Médica da Faculdade de Medicina da USP, sendo possível determinar dois parâmetros: a velocidade de propagação média entre eletrodos (cerca de 1.2 cm/s) e as componentes principais da freqüência basal (0.63 e 0.65 Hz). / The analysis of myoelectric signals from the gastro-intestinal tube of laboratory animals (mice), recorded by chronically implanted electrodes, is a key stone in understanding the disorders associated to the digestive system. This analysis meets considerable difficulties when done by classical methods, specially those based in the Fourier transform. The many myoelectric components interactions makes the following of these signals along the digestive tract and the retrieval of typical parameters (such as the propagation velocity between electrodes) a very complicated and laborious task. Here is studied a newer and more promising alternative: the Wavelet transform. Using this mathematical tool, it becomes possible to obtain a better time-frequency resolution of the studied signals, allowing to find patterns related to the signal propagation even in noisy and multifrequencial readings. 82 readings from 9 normal animals belonging to the Medical Investigation Laboratory of the Medicine Faculty of University of São Paulo were analyzed, becoming feasible to determine two parameters: the mean propagation velocity between electrodes (about 1.2 cm/s) and the main components of the basal frequency (0.63 e 0.65 Hz).

Dans le contexte du handicap, certaines solutions technologiques permettent de pallier des déficiences pour lesquelles la pharmacologie et la chirurgie sont impuissantes. Les neuroprothèses font partie de ces solutions technologiques. Il s'agit de dispositifs s'interfaçant avec le système nerveux (périphérique ou central), soit pour agir sur celui-ci (stimulation électrique fonctionnelle par exemple), soit pour y recueillir des signaux destinés à commander un dispositif extérieur tel qu'une prothèse robotisée.Le travail présenté dans ce manuscrit s'inscrit dans ce second contexte du recueil de signaux neuronaux sur un nerf périphérique. Aujourd'hui, le seul dispositif utilisable de manière chronique sur l'être humain est l'électrode Cuff tripolaire. Celle-ci recueille l'activité globale du nerf et manque singulièrement de sélectivité. Des dispositifs plus sélectifs, comme les électrodes intrafasciculaires, existent, mais présentent l'inconvénient d'être traumatisants pour le nerf et extrêmement délicats à mettre en place. Notre objectif, au cours de ce travail de thèse, a donc été de développer un dispositif associant la sélectivité d'une électrode intrafasciculaire à la faible invasivité d'une électrode Cuff.Nous avons donc commencé par étudier, en simulation, le potentiel d'action extracellulaire d'un axone myélinisé. Nos simulations nous ont permis de mettre en évidence un "phénomène local", évoqué dans plusieurs publications d'expérimentations antérieures à nos travaux et propre aux axones myélinisés, décroissant très rapidement quand augmente la distance de l'axone à l'électrode. Nous avons donc étudié et dimensionné un dispositif simple, spécifiquement sensible à ce phénomène local.A partir de ce dispositif, nous avons proposé une architecture d'électrode extraneurale qui possède un indice de sélectivité très supérieur à celui de la plus sélective des électrodes extraneurales publiées à ce jour.Malheureusement, ce gain en sélectivité se paie par une dégradation du rapport signal sur bruit. Nous avons donc étudié les solutions permettant de réduire le niveau de bruit ramené au niveau de l'électrode, et nous exposons les grandes lignes d'un dispositif électronique d'acquisition multi-voie faible bruit à température ambiante.Enfin, nous avons construit un modèle d'axone artificiel qui nous a permis de valider expérimentalement nos modèles de simulation ainsi que l'existence du phénomène local. / In the context of sensory or motor deficiencies, some technologic solutions can be proposed in case pharmacology and surgery are inefficient. Neuroprostheses are one of these technologic solutions. It consists in devices interfacing the (peripheral or central) nervous system, either acting on it (functional electric stimulation, neuromodulation...), or recording neuro-signals (automated prosthesis control, closed-loop stimulation...).The work presented in this manuscript focuses on the latter, and more precisely, on recording neuro-signals from peripheral nerves. Today, the only implantable device that can be used for chronic recording on human peripheral nervous system is the tripolar cuff electrode.Unfortunately, it is sensitive to the nerve global activity and exhibits a very low selectivity. More selective devices, like intrafascicular electrodes, exist, but has the drawback to be more traumatizing for the nerve.Therefore, the objective of this thesis was to develop a device associating the selectivity of intrafascicular electrode together with the low invasivity of cuff electrode.We thus started to perform simulations of the extracellular action potential of myelinated axons, putting in evidence a "local phenomenon", already described in some previous publications, and decreasing very quickly with the distance. Then, we have designed and studied a simple neural interface based on the characteristics of this local phenomenon, and specifically sensitive to it.The results have allowed us to propose an extraneural electrode, promising a selectivity index far higher than today's state of the art.Unfortunately, the gain in selectivity leads also to the degradation of signal-to-noise ratio. We have thus studied solutions to reduce noise at the electrode interface, and presented the architecture of a low-noise multi-channel acquisition circuit fitting our requirements.Finally, we have built the artificial model of an axon in order to experimentally validate simulation results, as well as the the local phenomenon characteristics.

The development of a cost-effective multiphase flow meter to determine the individual phase flow rates of oil, water and gas was investigated through the exploitation of a single clamp-on gamma densitometer and signal processing techniques. A fast-sampling (250 Hz) gamma densitometer was installed at the top of the 10.5 m high, 108.2 mm internal diameter, stainless steel catenary riser in the Cranfield University multiphase flow test facility. Gamma radiation attenuation data was collected for two photon energy ranges of the caesium-137 radioisotope based densitometer for a range of air, water and oil flow mixtures, spanning the facility’s delivery range. Signal analysis of the gamma densitometer data revealed the presence of quasi-periodic waveforms in the time-varying multiphase flow densities and discriminatory correlations between statistical features of the gamma count data and key multiphase flow parameters. The development of a mechanistic approach to infer the multiphase flow rates from the gamma attenuation information was investigated. A model for the determination of the individual phase flow rates was proposed based on the gamma attenuation levels; while quasi-periodic waveforms identified in the multiphase fluid density were observed to exhibit a strong correlation with the gas and liquid superficial phase velocity parameters at fixed water cuts. Analysis of the use of pattern recognition techniques to correlate the gamma densitometer data with the individual phase superficial velocities and the water cut was undertaken. Two neural network models were developed for comparison: a single multilayer-perceptron and a multilayer hierarchical flow regime dependent model. The pattern recognition systems were trained to map the temporal fluctuations in the multiphase mixture density with the individual phase flow rates using statistical features extracted from the gamma count signals as their inputs. Initial results yielded individual phase flow rate predictions to within ±10% based on flow regime specific correlations.

The development of a cost-effective multiphase flow meter to determine the individual phase flow rates of oil, water and gas was investigated through the exploitation of a single clamp-on gamma densitometer and signal processing techniques. A fast-sampling (250 Hz) gamma densitometer was installed at the top of the 10.5 m high, 108.2 mm internal diameter, stainless steel catenary riser in the Cranfield University multiphase flow test facility. Gamma radiation attenuation data was collected for two photon energy ranges of the caesium-137 radioisotope based densitometer for a range of air, water and oil flow mixtures, spanning the facility’s delivery range. Signal analysis of the gamma densitometer data revealed the presence of quasi-periodic waveforms in the time-varying multiphase flow densities and discriminatory correlations between statistical features of the gamma count data and key multiphase flow parameters. The development of a mechanistic approach to infer the multiphase flow rates from the gamma attenuation information was investigated. A model for the determination of the individual phase flow rates was proposed based on the gamma attenuation levels; while quasi-periodic waveforms identified in the multiphase fluid density were observed to exhibit a strong correlation with the gas and liquid superficial phase velocity parameters at fixed water cuts. Analysis of the use of pattern recognition techniques to correlate the gamma densitometer data with the individual phase superficial velocities and the water cut was undertaken. Two neural network models were developed for comparison: a single multilayer-perceptron and a multilayer hierarchical flow regime dependent model. The pattern recognition systems were trained to map the temporal fluctuations in the multiphase mixture density with the individual phase flow rates using statistical features extracted from the gamma count signals as their inputs. Initial results yielded individual phase flow rate predictions to within ±10% based on flow regime specific correlations.