The purpose of this study was to evaluate the accuracy in real time measurements acquired from GPS and GLONASS satellite observations using RTK techniques in an urban and forested environment. To determine this accuracy, 2 data sets of 3-dimensional coordinates were created and compared at 14 stations situated at East Tennessee State University. One data set included coordinates determined by conventional land survey methods; the second was solved by RTK GPS/GLONASS. Once the magnitude of any deviation in the coordinate positions was determined, the contributions to the accuracies from cycle slips, multipath, satellite availability, PDOP, and fixed or float solutions were evaluated. Three points in the urban environment varied from the conventional data set. Multipath was assumed to be the major bias in these points. Seven points in the forested environment varied from the conventional data set. The use of float solutions and high PDOP may have caused this bias.

Orientador: Daniele Barroca Marra Alves / Resumo: A observável fase da onda portadora é utilizada quando se deseja obter alta acurácia no posicionamento por sistemas globais de navegação por satélite. Mas, esta observável pode sofrer perdas de ciclos durante o rastreamento do sinal. Uma perda de ciclo é uma descontinuidade de um número inteiro de ciclos na fase da onda portadora, causada por uma perda temporária de rastreamento do sinal pelo receptor. Assim, as perdas de ciclos devem ser detectadas e corrigidas para que se obtenha posicionamento acurado. Diversos métodos foram desenvolvidos com esse objetivo. Algo que também vem sendo amplamente estudado é a detecção das perdas de ciclos nos novos sinais GPS. O advento da terceira frequência disponibilizada a partir da modernização do GPS também deve ser explorado. Na presente pesquisa buscou-se estudar os diferentes métodos de detecção e correção de perdas de ciclos e implementar o mais adequado para o posicionamento de alta acurácia no Brasil. Foram selecionadas duas metodologias para fim de detecção e correção, sendo a primeira baseada no método das triplas diferenças (TD) clássico e a segunda baseada no método de sequência de combinações lineares. Além disso, foi proposta uma terceira metodologia para melhoria do método das TD. Foram realizados dois experimentos, sendo o primeiro realizado com dados simulados e o segundo com dados reais em período de alta atividade ionosférica, e para ambos os casos foram considerados dois cenários: linha de base curta e linha de base lo... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Carrier phase observable are used in high accuracy positioning by global navigation satellite systems. However, the carrier phase measurement may suffer cycle slips during signal tracking. A cycle slip is a discontinuity of an integer number of cycles in the phase observable, caused by a temporary loss of lock in the receiver carrier tracking loop. Thus, cycle slips must be detected and corrected in order to obtain accurate positioning. Several methods were developed for this purpose. Something that has also been studied is the detection of cycle slip in the new GPS signals. The advent of the third frequency available from the GPS modernization. Should be explored in the context of cycle slip detection and correction. In the present research the different cycle slip detection and correction methods were studied and t the most suitable for the high accuracy positioning in Brazil was implemented. Two methodologies were selected for the purpose of this research, the first one based on the classical triple difference (TD) method and the second based on the linear combination sequence method. In addition, a third methodology was proposed to improve the TD method. Two experiments were performed, the first one being carried out with simulated data and the second with real data in the period of high ionospheric activity, and for both cases two scenarios were considered: short baseline and long baseline. For simulated short baseline data, the classical TD method, the proposed TD metho... (Complete abstract click electronic access below) / Mestre

Tandis que les humains explorent la nature sans relâche, ils font également attention à être plus conscients d'eux-mêmes, à mieux connaître ce qui les entourent, et, par exemple, à être informés de leurs positions, vitesses, ou trajectoires où qu'ils se trouvent. Les systèmes de positionnement par satellites (GNSS) ont fourni une manière de le faire à l'extérieur, devenant un assistant indispensable pour nous. Après le succès de GPS et GLONASS, Galileo et BeiDou sont actuellement en cours de déploiement, offrant plus de choix pour le positionnement autonome ou assisté. Toutefois, les signaux GNSS sont vulnérables aux obstacles: il n'y a presque pas de service GNSS à l'intérieur des bâtiments, dans les tunnels, ou dans les parkings souterrains; la continuité des services de positionnement par satellites n'est toujours pas assurée dans les "canyons urbains". Afin de faire face à ce problème, cette thèse est consacrée au positionnement en environnements contraints (où le GNSS n'est pas disponible), en y incluant la conception, les algorithmes et les technologies correspondantes. Un état de l'art est élaboré sur les systèmes de positionnement à radio ou à l'inertie. Parmi les technologies possibles, la discussion ne se limite pas à l'approche basée sur GNSS, mais elle est centrée sur cette dernière à cause de ses avantages et aussi de l’implication profonde de notre groupe de recherche dans ce domaine. Nous avons, d'une part, étudié la possibilité ainsi que les limites du positionnement avec une précision centimétrique en déployant notre système Repealite (i.e. un système de positionnement à l'intérieur d'une base des émetteurs GNSS spéciaux); et proposé, d'autre part, une méthode de localisation en groupe d'objets dynamiques communicants. Cette méthode procède d'une problématique délicate que l'on rencontre dans les approches GNSS, qu'est la détermination de la position initiale du récepteur. On montre en quoi elle permet également d'aller au-delà de la portée des approches classiques GNSS / While the human beings explore the nature tirelessly, they also put significant concerns to be aware of themselves, to know better of the circumstances, and to be informed with their precise positions, velocities, trajectories, and so on, in local environments. The Global Navigation Satellite Systems (GNSS) have provided an efficient method to do so outdoors, and have already become an indispensable assistant of many people. After the success of GPS and GLONASS, Galileo and BeiDou are currently under deployment, offering more choices of the independent or collaborative positioning. However, the GNSS signal is vulnerable to obstructions: almost no GNSS services are available inside buildings, tunnels, or underground parkings; the services are not always coherent in urban canyons. In order to address this problem, this thesis is dedicated to the design frameworks for the positioning in GNSS-challenged environments, as well as the corresponding algorithms and technologies. A brief survey of the latest radio-based and inertial positioning/tracking systems is provided. Among the feasible technologies, the discussion is centered on but not limited to the GNSS-based approach, which is due to the inherited advantages of this approach and also the deep engagement of our research group in this domain. We have, on one hand, explored the possibility and limitations on the centimeter-accuracy positioning with our Repealite system (i.e. a GNSS-base indoor positioning system with specific features); on the other hand, a method of batch localization for the nodes in a network of dynamic communicating objects is proposed, which is originated from an issue of the GNSS-based approach - the resolution of the receiver initial point, but then it goes beyond the scope of the “classical” GNSS-based approach

La thèse a pour but de développer des algorithmes robustes de poursuite de phase multifréquence en environnement dégradé. L’objectif est d’élaborer de nouvelles structures pouvant opérer à des niveaux de rapport signal à bruit inférieurs aux limites des algorithmes actuellement implémentés dans des récepteurs grand public. Les problèmes de robustesse des algorithmes d’estimation de phase étant en grande partie causés par le phénomène de sauts de cycle, les différents axes de recherche se sont focalisés sur des nouvelles approches de développement de phase au sein des structures de poursuite. Pour ce faire, deux approches ont été étudiées et testées. Dans un premier temps, deux structures de poursuite monofréquence basées sur une DPLL conventionnelle ont été développées. Ces structures disposent d’un système externe de développement de phase visant à prédire et pré-compenser la sortie du discriminateur grâce à l’analyse des sorties du discriminateur ou des sorties du filtre de boucle. La réduction de la dynamique à estimer va alors permettre de réduire l’apparition des sauts de cycle se produisant au niveau du discriminateur. Par la suite, ce système de développement de phase a été adapté à la poursuite de phase multifréquence. Grâce à l’exploitation de la diversité en fréquence offerte par les signaux de navigation (i.e., de la proportionnalité des fréquences Doppler), il a été possible de mettre en place une étape de fusion de données qui a permis d’améliorer la précision de la prédiction de la sortie du discriminateur et donc d’améliorer la robustesse de la structure. Dans un second temps, les travaux de recherche se son taxés sur une nouvelle approche de poursuite de phase et de correction du phénomène de sauts de cycle basée sur une technique de filtrage Bayésien variationnel. Toujours en exploitant la diversité en fréquence des signaux de navigation, cette méthode suppose un modèle de dynamique de phase Markovien qui va imposer une certaine continuité de l’estimation et va permettre de fournir une estimation de phase développée. / This thesis aims at introducing multifrequency phase tracking algorithms operating in low C/N0environment. The objective is to develop new structures whose tracking limits are lower than thatof current algorithms used in mass market receivers. Phase tracking suffers from a lack of robustnessdue to the cycle slip phenomenon. Works have thus been focused on elaborating new phaseunwrapping systems. To do so, two different tracking approaches were studied. First, we have developed new monofrequency tracking loops based on a conventional DPLL. These structures aimat predicting the discriminator output by analyzing, thanks to a polynomial model, the last outputsamples of either the discriminator or the loop filter. Once the discriminator output is predicted,the estimated value is pre-compensated so that the phase dynamics to be tracked is reduced aswell as the cycle slip rate. Then, the unwrapping structure analyzing the loop filter outputs hasbeen extended to multifrequency signals. Using a data fusion step, the new multifrequency structuretakes advantage of the frequency diversity of a GNSS signal (i.e., proportionality of Dopplerfrequencies) to improve the tracking performances. Secondly, studies have been focused on developing a new multifrequency tracking algorithm using variational Bayesian filtering technique.This tracking method, which also uses the GNSS frequency diversity, assumes a Markovian phasedynamics that enforces the smoothness of the phase estimation and unwraps it.