[pt] Radares são sistemas eletromagnéticos de detecção e localização de objetos refletores. Sua operação consiste na transmissão de formas de onda e na detecção e processamento do sinal refletido no alvo, visando extrair algumas características deste (e.g. posição, velocidade, direção, tipo de alvo). Radares de ruído, por sua vez, são sistemas que transmitem sinais com forma de onda aleatória e que, por este motivo, apresentam elevado desempenho em presença de interferências externas, intencionais ou não, e supressão da ambiguidade em distância na detecção de alvos. O presente trabalho desenvolve inicialmente uma modelagem probabilística dos sinais envolvidos em radares de ruído. Esta modelagem permite a obtenção de expressões fechadas para funções autocorrelação, densidades espectrais de potência, funções correlação cruzada e ganhos de processamento associadas aos diversos sinais envolvidos. Com base nesta modelagem, e feita uma análise comparativa de radares que utilizam compressão de pulso e integração em sua cadeia de recepção. Mais especificamente, são comparados radares de ruído com radares tradicionais (que utilizam formas de onda determinísticas - chirps) operando em cenários que permitem avaliar aspectos como a resolucao em distância, o nível de lóbulos secundários e a eliminação de ambiguidades. / [en] Radars are electromagnetics systems used for detecting and locating reflecting objects. Their operation is based on the transmition of waveforms and on the detection and processing of the reflected signal so that some targets can be estimated (e.g. position, velocity, direction and type of target). Noise radars are systems that use random signals as the transmitting waveform and that for this reason present high performance against external interference, intentional or not, and the suppression of range ambiguity in the detection of targets. The present work initially develops a probabilistic model for the random signals involved in Noise Radars. This model allows for derivation of closed forms expressions for autocorrelation functions, power spectral densities, processing gains, and cross-correlation functions of the random signals involved. Based on this model, a comparison of radars that employ pulse compression and integration in their receivers. More specifically, Noise Radars are compared to traditional radars that use deterministic waveforms (chirps) under different scenarios which are chosen so that some specific aspects like range resolution, side-lobe levels and range ambiguity elimination can be evaluated.

Non-invasive and fast flow measurement techniques have had increasing importance for the last decades. Scientists are looking for such quick techniques to be able to monitor real velocities without disturbing flow itself. Ultrasound Doppler velocimetry (UDV) being one of such techniques promising with advantages of getting simultaneous velocity measurements from several points and of applicability for opaque liquids as well. UDV is a technique which is still being developed for new applications and analysis of complex flows. In this study effect of sinusoidal oscillating, turbulent (random) and viscoelastic fluid motions on UDV signals were investigated theoretically and experimentally. Obtained mathematical relations for random and viscoelastic motions were utilized to get statistics of flow and distribution of relaxation spectrum, respectively. Analytical analysis and numerical simulation of sinusoidal oscillating flow depicted that there is a critical value for the ratio of oscillation amplitude to oscillation frequency for a specified set of measurement parameters of UDV. Above this critical value UDV is not successful to determine mean flow velocity. Mathematical relations between velocity probability density function (PDF) &ndash / velocity auto correlation function (ACF) and UDV signal spectrum were obtained in the analysis v of flow with random velocity. Comparison of velocity ACFs from direct velocity measurements and from raw in-phase (I) and quadrature (Q) signals through derived relation, revealed that time resolution of UDV technique is not enough for getting a good velocity ACF and thus turbulence spectrum. Using I and Q signals rather than measured velocities to get velocity ACF, increased the time resolution in the order of number of pulses used for getting one velocity value (Nprn). Velocity PDF obtained from UDV spectrum was compared with the one obtained from measured velocities with the assumption of Gaussian PDF. Both velocity PDFs were consistent. Also some parameters of pipe turbulence from literature were compared with the presented findings from velocity ACF obtained from I and Q signals through derived relation. Results showed good compatibility. In the last part of the study, complex viscosity of a linear viscoelastic fluid mathematically related to spectrum of UDV for a pipe flow with small-amplitude oscillating pressure field. Generalized Maxwell model was employed to express complex viscosity terms. Zero frequency (mean flow) component of UDV spectrum was used to obtain an equation for relaxation viscosities of generalized Maxwell model. Results have revealed that UDV technique can also be used to probe some of viscoelastic material functions. In conclusion, UDV is relatively new but a promising technique for the measurement and analysis of complex flows in a non-invasive manner.

Zur Verbesserung der Störsicherheit bei der digitalen Signalübertragung wird eine Hysterese in die Übertragungskennlinie des Gatters eingefügt. Der Einfluss der Höhe der beiden Schwellwerte auf die Anzahl der auftretenden Fehler wird mittels eines Rechnerprogrammes untersucht. Ein Zufallsgenerator erzeugt Signale in verschiedenen Höhen und Breiten, die sich den ungestörten Signalen überlagern. Es erfolgt eine Umsetzung einer integrierten Schaltung auf einem TTL Master. Die Schaltung wird mittels eines eigens entwickelten Netzwerkanalyseprgrammes berechnet. Messergebnisse werden mitgeteilt. / To enhance the noise immunity of digital signal transmission, a hysteresis is introduced to the transfer characteristic of integrated digital circuit. The influence of height of the two threshold values to the number of occurring errors is examined by a computer program. A random number generator generates signals of different heights and widths, which are superimposed on the undisturbed signals. There is an implementation of an integrated circuit on a TTL master. The DC performane is calculated by means of a specially developed circuit analysis program. Measurement results are presented.

Le travail réalisé dans le cadre de cette thèse a pour but de caractériser des signaux aléatoires, rencontrés dans le domaine aérien et sous-marin, en s’appuyant sur une approche statistique. En traitement du signal, l'analyse statistique a longtemps été fondée sous l'hypothèse de Gaussianité des données. Cependant, ce modèle n'est plus valide dès lors que la densité de probabilité des données se caractérise par des phénomènes de queues lourdes et d'asymétrie. Une famille de lois est particulièrement adaptée pour représenter de tels phénomènes : les distributions α-stables. Dans un premier temps, les distributions α-stables ont été présentées et utilisées pour estimer des données synthétiques et réelles, issues d'un sondeur monofaisceau, dans une stratégie de classification de fonds marins. La classification est réalisée à partir de la théorie des fonctions de croyance, permettant ainsi de prendre en compte l'imprécision et l'incertitude liées aux données et à l'estimation de celles-ci. Les résultats obtenus ont été comparés à un classifieur Bayésien. Dans un second temps, dans le contexte de la surveillance maritime, une approche statistique à partir des distributions α-stables a été réalisée afin de caractériser les échos indésirables réfléchis par la surface maritime, appelés aussi fouillis de mer, où la surface est observée en configuration bistatique. La surface maritime a d'abord été générée à partir du spectre d'Elfouhaily puis la Surface Équivalente Radar (SER) de celle-ci a été déterminée à partir de l'Optique Physique (OP). Les distributions de Weibull et ont été utilisées et comparées au modèle α-stable. La validité de chaque modèle a été étudiée à partir d'un test de Kolmogorov-Smirnov. / The purpose of this thesis is to characterize random signals, observed in air and underwater context, by using a statistical approach. In signal processing, the hypothesis of Gaussian model is often used for a statistical study. However, the Gaussian model is not valid when the probability density function of data are characterized by heavy-tailed and skewness phenomena. A family of laws can fit these phenomena: the α-stable distributions. Firstly, the α-stable distribution have been used to estimate generated and real data, extracted from a mono-beam echo-sounder, for seabed sediments classification. The classification is made by using the theory of belief functions, which can take into account the imprecision and uncertainty of data and theirs estimations. The results have been compared to a Bayesian approach. Secondly, in a context a marine surveillance, a statistical study from the α-stable distribution has been made to characterize undesirable echo reflected by a sea surface, called sea clutter, where the sea surface is considered in a bistatic configuration. The sea surface has been firstly generated by the Elfouhaily sea spectrum and the Radar Cross Section (RCS) of the sea surface has been computed by the Physical Optics (PO). The Weibull and distributions have been used and the results compared to the α-stable model. The validity of each model has been evaluated by a Kolmogorov-Smirnov test.

Zur Verbesserung der Störsicherheit bei der digitalen Signalübertragung wird eine Hysterese in die Übertragungskennlinie des Gatters eingefügt. Der Einfluss der Höhe der beiden Schwellwerte auf die Anzahl der auftretenden Fehler wird mittels eines Rechnerprogrammes untersucht. Ein Zufallsgenerator erzeugt Signale in verschiedenen Höhen und Breiten, die sich den ungestörten Signalen überlagern. Es erfolgt eine Umsetzung einer integrierten Schaltung auf einem TTL Master. Die Schaltung wird mittels eines eigens entwickelten Netzwerkanalyseprgrammes berechnet. Messergebnisse werden mitgeteilt. / To enhance the noise immunity of digital signal transmission, a hysteresis is introduced to the transfer characteristic of integrated digital circuit. The influence of height of the two threshold values to the number of occurring errors is examined by a computer program. A random number generator generates signals of different heights and widths, which are superimposed on the undisturbed signals. There is an implementation of an integrated circuit on a TTL master. The DC performane is calculated by means of a specially developed circuit analysis program. Measurement results are presented.