Este trabalho concentra-se no aumento da robustez à taxa de erros em redes de sensores sem fio industriais (RSSFI). Para alcançar tal objetivo, é preciso, de algum modo, contornar os problemas intrínsecos da comunicação sem fio, que são, o ruído ambiental, interferências e desvanecimento por multipercurso. O multipercurso pode ser considerado um dos principais fatores que tornam as comunicações nas redes de sensores sem fio (RSSF) um verdadeiro desafio quando comparadas com outros meios, como a transmissão em fibra óptica, cabo ou mesmo transmissões de rádio ponto-a-ponto. Nesse sentido, propõe-se a diversidade de antenas como uma solução para minimizar os efeitos do multicaminho, com o objetivo de melhorar a confiabilidade do enlace de rádio para permitir o emprego de RSSF densas. O potencial da diversidade de antenas em RSSF não está totalmente explorado em aplicações industriais. Esta dissertação apresenta ainda, a técnica “Combinação de seleção” para RSSFI através de um algoritmo que seleciona a porta do receptor que apresenta o melhor indicador de qualidade de enlace e realiza a comutação das antenas nos módulos de rádio. Os resultados são analisados para dois tipos de enlaces, isto é, com e sem diversidade de antenas na recepção, e discute-se opções para melhorar o PER (“Packet Error Rate”) com as atuais técnicas de diversidade. / This work concentrates on the increase of reliability and robustness in Industrial Wireless Sensor Networks (IWSNs), decreasing the Packet Error Rate (PER). To achieve this objective, is need to somehow circumvent and reduce the underlying problems of wireless communication, which are: environmental noise, interference, and multipath fading. The multipath can be seen as the main factor which becomes the communications in the Wireless Sensor Networks (WSNs) a real challenge when compared to other types of means such as a transmission in fiber, cable or even point-to-point radio transmissions. In this sense, it is proposed antenna diversity as solution to reduce these effects, with the aim to improve the reliability of the radio link to allow the use of dense WSNs. The potential of antenna diversity in WSN is not fully exploited in industrial applications. In this sense, this dissertation presents the "Combination of selection" technique for IWSNs through an internal algorithm that selects the receiver port that has the highest LQI ("Link Quality Indicator") and performs an antenna switching in the radio modules. Also, options to improve PER using diversity techniques are discussed.

[pt] Sistemas com múltiplas antenas transmissoras e receptoras, também conhecidos como sistemas MIMO (Multiple Input-Multiple Output), têm sido apontados como uma solução para aumentar a capacidade de enlaces sem fio, permitindo aos usuários utilizar aplicações com altas taxas de dados. Isto é extremamente importante em sistemas onde a capacidade obtida com as técnicas tradicionais é bastante limitada devido às características do ambiente de propagação. Com o sistema MIMO, algumas destas características são exploradas para criar canais paralelos e obter aumento expressivo de capacidade. A análise da capacidade de sistemas MIMO se baseia em uma modelagem desenvolvida a partir do comportamento estatístico dos pares de enlaces existentes entre as múltiplas antenas transmissoras e receptoras. Existe, portanto, um grande interesse em medir este comportamento para situações típicas bem como em relacioná-lo a determinados parâmetros do sistema. Nesta tese apresentam-se os resultados de uma campanha de medidas visando caracterizar canais MIMO de faixa estreita e faixa larga em ambientes fechados (indoor) com uma freqüência de portadora de 2GHz. A partir dos dados medidos, avalia-se a capacidade e diversos parâmetros do canal espaço-temporal. Os parâmetros do canal MIMO são estimados através do algoritmo FD-SAGE e as dispersões temporal e espacial do canal são calculadas a partir dos parâmetros estimados. Uma análise dos autovalores da matriz do canal MIMO é realizada com o objetivo de relacionar os valores da capacidade ao número de canais paralelos. É analisada, ainda, a correlação entre a capacidade e os parâmetros físicos do canal, tais como espaçamento entre os elementos do arranjo, espalhamento angular, espalhamento dos retardos, número e potência dos multipercursos. / [en] Multiple antenna systems known as MIMO (Multiple Input Multiple Output) systems have been proposed as an effective way to address the user demand for high data rate applications in wireless systems. This is especially important in systems where the capacity attained with traditional techniques is very limited due to the adverse characteristics of the propagation environment. With MIMO, some of these characteristics are used to create parallel channels producing significant increase in capacity. The analysis of MIMO capacity is based on models developed from the statistical behavior of the multiple links between the transmitting and receiving antennas, and therefore there has been large interest in measuring these characteristics in typical scenarios and in relating the data to system parameters. In this thesis the results of a MIMO wideband measurement campaign carried out in an indoor scenario with a carrier frequency of 2 GHz is presented. The wideband and narrowband channel capacity and several channel parameters are evaluated from the measured data. The channel parameters are estimated using the frequency domain Space-Alternating Generalized Expectation maximization (FD- SAGE) algorithm. Temporal and spatial dispersions of the multipath channel are calculated from the estimated parameters and an eigenvalue analysis is performed seeking to relate the capacity values to the number of parallel channels. In addition, the correlation between channel capacity and physical parameters as antenna spacing, angle spread, delay spread, number and power of multipath components is investigated.

Dans les systèmes de positionnement par satellite (GNSS), les réflexions multiples, caractéristiques des canaux de propagation urbains, posent de gros problèmes dans la bonne estimation de la position. Bien que de nombreuses solutions aient été proposées pour lutter contre les multi-trajets (Narrow correlator, MEDLL), les multi-trajets à faibles retards relatifs (<0.1Chips) sont toujours problématique. Plus récemment, l'utilisation de réseaux d'antennes adaptatifs a été proposée pour lutter contre les multi-trajets. En effet, l'échantillonnage spatial du front d'onde réalisé à partir de plusieurs antennes permet de discriminer les sources dans le domaine spatial, et cela quels que soient les retards des multi-trajets. Cette thèse a donc pour objectif de définir des méthodes de réjection des multi-trajets qui tirent partie de la dimension des directions d'arrivées (DOA) apportée par un réseau d'antenne. Le cahier des charges impose que le réseau utilisé soit de petite taille (typiquement réseau carré 2×2), et demande un algorithme robuste aux défauts technologiques.Inspiré des méthodes de réjection d'interférence, les premières solutions proposées ont été les techniques d'antennes adaptatives afin de filtrer spatialement les multi-trajets Cependant, en raison de la petite taille du réseauainsi que de la corrélation intrinsèque entre les multi-trajets et le signal direct, ces méthodes ont rapidement été mises en défaut. En particulier, les multi-trajets spatialement proches du trajet direct restent toujours problématiques.Afin de tirer partie de la dimension spatiale apportée par un réseau d'antennes et sans être trop dépendant de l'espacement angulaire entre les sources, nous avons choisi de joindre à l'estimation des DOA, l'estimation des retards et fréquences Doppler de chaque trajet reçu. L'algorithme SAGE, issu de la théorie du maximum de vraisemblance, a été utilisé afin d'estimer de façon jointe les paramètres des sources. De plus, nous avons proposé une nouvelle implémentation de SAGE qui permet de réduire d'un facteur 500 la complexité de l'algorithme tout en conservant les mêmes performances d'estimation. Les simulations ont montré une nette amélioration de la réjection des multi-trajets par rapport aux méthodes mono-antenne et multi-antennes adaptatives, tout en conservant une complexité calculatoire raisonnable. Cette méthode a fait l’objet d’un brevet.Nous avons ensuite étudié l'influence des défauts technologiques (couplage, défaut de chaîne RF …), numériques (quantification) et des défauts intrinsèques à l'algorithme SAGE (estimation du nombre de multi-trajets) sur les performances d'estimation, et proposé différentes méthodes de compensation. Nous retiendrons que des filtres FIR ont été utilisés pour compenser les défauts large bande de la chaîne RF, et que le couplage peut être estimé à partir de la connaissance des DOA des signaux utiles. Les simulations ont montré qu'avec ces méthodes de compensation, l'algorithme SAGE affichait des performances très proches de celles dans le cas parfait (sans défaut). Un algorithme d'estimation du nombre de multi-trajets a aussi été proposé, et les résultats en simulation dynamique (en utilisant des modèles de canal existants) ont montré une bonne adaptation aux différentes situations.Pour finir, il est important de noter que les outils développés durant cette thèse peuvent être utilisés dans le cadre de l'étude du canal de propagation des signaux GNSS, en identifiant et estimant les multi-trajets susceptibles de perturber l'estimation de la position du récepteur. / In Global Navigation Satellite Systems (GNSS) applications, multipath (MP) errors are still one of the major error sources in conventional receivers. The additional signal replicas due to reflections introduce a bias in Delay Lock Loops (DLL), which finally leads to a positioning error. Several techniques have been developed for multipath mitigation or estimation such as the Narrow Correlator Spacing or the Multipath Estimating Delay-Lock-Loop (MEDLL) algorithm.However, these techniques suffers from high sensitivity to noise, and can not mitigate short delay multipath (<0.1 chip).More recently, the use of antenna array algorithms has been proposed for multipath mitigation. Antenna arrays perform a spatial sampling that makes possible the discrimination of sources in the space domain (azimuth and elevation).However, in conventional receivers, little room remains for antenna integration, and only a small number of antenna elements can be integrated. This study will therefore focus on algorithms for a 2x2 square antenna array. Moreover, theproposed solutions have to be robust against technological defects.Two solutions are investigated to mitigate multipath with an antenna array. The first one tries to filter the multipaths in the space domain in order to "clean" the incoming signal of all the multipaths. However, the results obtained with this solution are quite mitigated. Indeed, the small size of the array implies a low space resolution, and the correlation between the LOS signal and the multipaths strongly degrade the performance of high resolution algorithms. Thus, close spaced multipath are still a problem.In the second approach, a set of parameters (amplitudes, times-delays, Doppler shifts, elevations and azimuths) of all the incoming sources are estimated. The main difference with the first approach is that, instead of filtering the sources on the space domain only, the different incoming paths are filtered on space, time and frequency domains. To estimate the parameters of all the sources, SAGE algorithm, which is a low-complexity generalization of maximum likelihood theory, has been considered. Moreover, a new implementation of the SAGE algorithm has been investigated in order to reduce the complexity by a factor 500, without loss of estimation performances. The simulations show a real improvement in the multipath mitigation compared to mono antenna algorithms and beamformerapproaches.The impact of technological defects (mutual coupling, RF channel mismatch …), numerical defect (quantization) and SAGE defect (estimation of the number of path) on the estimation performances were also investigated, and severalcompensation algorithms were proposed. The wide band effects of the RF filter were compensated by FIR equalizer, and mutual coupling can be estimated thanks to the knowledge of the satellites DOA. Simulations show that the estimationperformance of the SAGE algorithm after array calibration are very close than the performance in perfect system. Last, we proposed an algorithm to estimate the number of path, and dynamic simulations (by using channel model) show avery good adaptation of the algorithm.Last but not least, the tools developed in this PhD can be also useful in multipath modelling applications for GNSS.

Étant donné l'urgence environnementale, le développement des énergies renouvelables s'est fortement accru ces dernières années. L'implantation de champs d'éoliennes est notamment en pleine expansion dans toute l'Europe. Ces éoliennes, de structure diélectrique et métallique et de grande taille, peuvent avoir un impact significatif sur les systèmes radiofréquences. En particulier, les systèmes de radionavigation et de surveillance opérés par la DGAC (VOR, radar) doivent cohabiter avec de nouveaux champs d'éoliennes. En effet, ces dernières influent sur le champ électromagnétique des systèmes et peuvent dégrader leurs performances (multitrajets, masquages... ). Il est alors important de pouvoir quantifier ces dégradations, en particulier dans le cas du VOR où les multitrajets engendrent une erreur de relèvement. Dans ce travail de thèse, un simulateur électromagnétique appelé VERSO (VOR ERror SimulatOr) est développé. Il permet d'estimer l'impact d'objets diffractants, en particulier d'éoliennes, sur le signal VOR. Dans la littérature, différentes techniques de modélisation sont proposées pour prédire ces phénomènes. Certaines sont trop approximatives, d'autres trop coûteuses en temps. Ainsi, le choix des techniques utilisées dans ce simulateur a été guidé par le compromis entre précision et temps de calcul. L'équation parabolique est utilisée pour modéliser la propagation de la source jusqu'aux éoliennes afin de prendre en compte le relief. Ensuite, une méthode d'optique physique sur matériaux diélectriques est mise en œuvre pour calculer le champ diffracté par ces objets. Le modèle électromagnétique de l'éolienne et les hypothèses inhérentes aux méthodes utilisées par VERSO ont été validés aux fréquences VHF (VOR) par comparaison avec la méthode des moments qui fait office de référence. Une extension de VERSO pour les systèmes radars a été abordée. Par conséquent des validations similaires à ces fréquences ont été réalisées. Plus spécifiquement, un modèle de pale d'éolienne prenant en compte la présence du parafoudre est proposé aux fréquences VOR et radar. Les effets de masquage sont eux aussi quantifiés pour ces 2 domaines de fréquences. Il est notamment démontré que l'effet de masquage pour des éoliennes alignées radialement à un VOR est négligeable. Cette approximation ainsi que le modèle de pale sont ensuite utilisés dans le programme VERSO. Ce dernier est validé à l'aide de contrôles en vol sur un scénario de 9 éoliennes implantées à 5 km du VOR de Boulogne-sur-Mer. Une étude quantitative de l'impact de chaque partie des éoliennes est menée afin de discriminer la source majoritaire d'erreur VOR. On constate notamment qu'à grande distance du VOR dans le scénario d'observation considéré, le mât constitue le principal contributeur en terme de champ diffracté et d'erreur VOR. Enfin, une étude statistique sur l'erreur VOR a permis d'obtenir un simulateur de scénarios qui donne l'erreur maximale avec une confiance fixée, en minimisant le nombre de simulations à effectuer. Pour accélérer cette méthode, une expression analytique approchée de l'erreur VOR maximum a été développée en fonction de la distance d'implantation et de la hauteur du mât de l'éolienne. Cette dernière étude fournit une méthode rapide pour évaluer l'impact de la construction d'un champ d'éoliennes quelconque à proximité de systèmes de l'aviation civile. / Considering the ecological emergency, the renewable energy development has greatly increased for a decade. In particular, the windfarms implantation rapidly expands in Europe. These windturbines are large obstacles composed by dielectric and metallic materials. So their impact on electromagnetic devices is significant. The radionavigation systems for the civil aviation services are particularly concerned. However, they have to work side with new windfarms. Actually, these latter cause scattering effects on the electromagnetic signals and can degrade the performances of these equipments (multipaths, shadowing effects etc.). Thus, quantifying these degradations is crucial, particularly on the VOR devices where multipath effects cause an error on the azimuth. In this thesis work, an electromagnetic simulator called VERSO (VOR ERror SimulatOr) is developed. It can estimate the impact of scattering objects, especially windturbines, on the VOR signal. In literature, several techniques are proposed to model these phenomena: some make coarse approximations and some others are memory intensive. Thus, the choice over the methods used in VERSO is a compromise between precision and memory requirement. The parabolic equation is used to model the propagation from the source to the windturbines so as to take the relief into account. A physical optic based method is used to compute the field scattered by these objects. The electromagnetic model of the windturbine and the hypothesis due to the methods used by VERSO have been validated in the VHF (VOR) frequency by comparison with the method of moments, which is the reference. An extension of VERSO for the radar systems is introduced. Consequently similar validations have been performed at radar frequencies. In particular, a windturbine blade model taking into account the lightning protection is proposed for the VOR and the radar frequencies. The shadowing effects are also quantified in both frequency domains. Especially, a demonstration that the shadowing effects due to radially implemented windturbines can be neglected around a VOR beacon is proposed. This approximation and the blade model are used for the implementation of VERSO. This simulator is validated by comparison with measurements on 9 windturbines built 5~km far from a VORC in Boulogne-sur-Mer (France). A study is performed to quantify the influence of each part of the windturbine. The mast is shown to be the main contributor regarding to the electromagnetic field and the VOR error. Finally, parametric simulations are performed and analytic expressions are proposed to describe the evolution of the maximum VOR error with respect to the mast size and the distance VOR-windturbine. The latter study gives some key parameters that need to be considered for the elaboration of a windfarm building plan close to civil aviation systems for the project to be viable.

The purpose of this work was investigate and measure the possibility to exist multipath caused by the corona noise generated in high voltage electric lines recepting GNSS signals. Experiments were realized with a 69 kV tension electric line localized in Santa Maria/RS. A serie was located with eleven transversal points to the longitudinal ranging of the transmission line. In this points, the eletromagnetic field was estimated using an equipment to measure the electromagnetic field EM-8000 model. Valuations was realized about the interferences on positioning by the absolut method, using a GNSS receiver with C/A code Garmin GPS II Plus model and the positioning by the relative method with the GNSS receiver L1/L2 Topcon Hiper model. To estimate the interference in the points was necessary to obtain the geodesic coordinates of this points without multipath possibilities caused of corona noise. For this was implanted a squared topographical polygonal where the support points were located far from the transmission lines, free of the electromagnetic field influence. Calculating the variances among the coordinates was necessary transform the geodesic coordinates obtained at GNSS receivers to the Local Topographic System, to make the systems compatible, using the rotation and translation model. Calculating was possible conclude that to the raising method and the geodesic equipment used in this experiment, in a 69 kV transmission line, the found discrepancies were in the standard deviation of 17,10 mm admitted to equipment GNSS receiver L1/L2 and 15 mm to GNSS receiver C/A code, in a 95% confidence level. So it was not evidenced interferences in the electromagnetic field at GNSS signal reception. / O propósito deste trabalho foi investigar e mensurar a possibilidade de haver multicaminhamento pelo efeito corona gerado em linhas elétricas de alta tensão na recepção de sinais GNSS. Foram realizados experimentos em uma rede elétrica com tensão de 69 kV localizado no município de Santa Maria/RS. Foi implantada uma série de onze pontos transversais ao alinhamento longitudinal da linha de transmissão. Nestes pontos realizou-se uma avaliação do campo eletromagnético gerado utilizando um equipamento medidor de campo eletromagnético modelo EM- 8000. Foram realizadas avaliações da interferência no posicionamento pelo método absoluto utilizando o receptor GNSS de código C/A Garmin modelo GPS II Plus e o posicionamento pelo método relativo com o receptor GNSS portadora L1/L2 Topcon modelo Hiper. Para avaliar a interferência dos pontos foi necessário obter as coordenadas geodésicas desses pontos sem possibilidades de multicaminhamento pelo efeito corona. Para isto implantou-se uma poligonal topográfica enquadrada em que os marcos de apoio localizavam-se distantes a linha de transmissão sem a influência do campo eletromagnético. Ao realizar o cálculo das discrepâncias entre as coordenadas foi necessário realizar a transformação das coordenadas geodésicas obtidas pelos receptores GNSS para o Sistema Topográfico Local para a compatibilização dos sistemas, sendo utilizado o modelo das rotações e translações. Realizando os cálculos concluiu-se que para os métodos de levantamento e os equipamentos geodésicos utilizados neste experimento, em uma linha de transmissão de 69 kV, as discrepâncias constatadas ficaram dentro do desviospadrões de 17,10 mm admitido ao equipamento receptor GNSS portadora L1/L2 e 15 m ao receptor GNSS código C/A, a nível de confiança de 95%. Assim não se constatou interferências do campo eletromagnético na recepção dos sinais GNSS.

Submitted by Fernando Souza (fernandoafsou@gmail.com) on 2017-08-14T10:56:32Z No. of bitstreams: 1 arquivototal.pdf: 4889368 bytes, checksum: 2154a6772c98f0f2fb36fa3148dc4a24 (MD5) / Made available in DSpace on 2017-08-14T10:56:32Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 4889368 bytes, checksum: 2154a6772c98f0f2fb36fa3148dc4a24 (MD5) Previous issue date: 2016-02-29 / A Wireless Sensor Network (WSN) can be characterized by the use of sensor nodes in varied environments for many different applications, e.g., monitoring and control systems applications in industrial environments. The use of WSNs in industrial applications must deal with typical problems of wireless networks, such as high attenuation and obstructions. In these environments, different characteristics of the channels can be observed due to its multipath profile. Changes in the topology may also cause changes in the channel characteristics as the time goes on, and it may cause differences in the average value of the received power. An alternative to deal with it and with the variation of channel quality is the development of protocols that use multiple channels. Its use allows, e.g., to achieve better use of the channel, increasing the network transmission capacity. However, there is a difficulty in evaluating and comparing different approaches due to the lack of simulators which have more accurate models. To assess the real benefits of these protocols, one should take into account a more realistic propagation model. This work presents a realistic simulation model that captures the effects of fading on a large and small scale, log-normal shadowing, the non-stationary characteristics of the channel for a long time, and the different characteristics of the different IEEE 802.15.4 standard channels. The model was then built into the open-source discrete event simulator Castalia and tested, comparing it with the default stationary model of the simulator. After integration, the protocol Tree-based Multi-Channel Protocol (TMCP) was evaluated using the proposed model. In the experiments it was possible to identify good accuracy in both the model and the TMCP protocol. / Uma Rede de Sensores sem Fio (RSSF) pode ser caracterizada pela utilização de nós sensores em variados ambientes para as mais diversas aplicações, por exemplo, para monitoramento e controle em ambiente industrial. O uso de RSSFs para implementar aplicações industriais precisa lidar com problemas típicos das redes sem fio, como alta atenuação e obstruções. Nesses ambientes, diferenças nas características dos canais podem ser observadas dado o seu perfil de multipercurso. Mudanças na topologia também podem provocar mudanças nas características dos canais ao longo do tempo, podendo causar diferença no valor médio da potência recebida. Uma alternativa para lidar com isso e com a variação da qualidade dos canais é o desenvolvimento de protocolos que utilizam múltiplos canais. Seu uso permite, por exemplo, alcançar uma melhor utilização do canal, aumentando a capacidade de transmissão da rede. No entanto, existe uma dificuldade em avaliar e comparar as diferentes abordagens devido à carência de simuladores que apresentem modelos mais precisos. Para avaliar os reais benefícios desses protocolos deve-se levar em consideração um modelo de propagação mais próximo da realidade. Esse trabalho apresenta um modelo de simulação que captura os efeitos de desvanecimento em larga escala, atenuação em pequena escala, sombreamento log-normal, as características não-estacionárias do canal em longos períodos de tempo, e as características distintas dos diferentes canais do padrão IEEE 802.15.4. O modelo foi então integrado ao simulador de eventos discretos de código aberto Castalia e testado, comparando-o com o modelo estacionário padrão do próprio simulador. A partir da integração, o protocolo Tree-based Multi-Channel Protocol (TMCP) foi implementado e avaliado utilizando o modelo proposto. Com os experimentos foi possível identificar uma ótima precisão do modelo e do protocolo TMCP.

Abstract This dissertation presents complete parameterizations for a three-dimensional (3-D) geometry-based stochastic radio channel model (GSCM) at 10 GHz based on measurement campaigns. The thesis is divided into the following main parts: radio channel measurements, the characterization of model parameters, and model validation. Experimental multiple-input multiple-output (MIMO) channel measurements carried out in two-story lobby and urban small cell scenarios are first described. The measurements were performed with a vector network analyzer and dual polarized virtual antenna arrays with a bandwidth over 500 MHz. The measurement data was post-processed using the ESPRIT algorithm and the post-processed data was verified using a semi-deterministic map-based model. The results showed a good match between estimated and modeled multipath components (MPCs). In addition, single-input single-output outdoor-to-indoor measurements were executed through a standard multi-pane glass window and concrete wall. A statistical analysis was carried out for defining full 3-D characterization of the propagation channel in both line-of-sight (LOS) and non-line-of-sight (NLOS) propagation conditions. The delay and angular dispersions of MPCs are smaller in comparison to lower frequency bands due to the higher attenuation of the delayed MPCs. Moreover, specular reflection is observed to be the more dominant propagation mechanism in comparison to diffuse scattering, leading to smaller cluster angle spreads in comparison to lower frequency bands. The penetration loss caused by a standard multi-pane glass window is on the same level as in the lower frequency bands, whereas the loss caused by the concrete wall is a few dBs higher than at lower frequency bands. Finally, the GSCM with determined parameters is validated. A MIMO channel was reconstructed by embedding 3-D radiation patterns of the antennas into the propagation path estimates. Equivalently the channel simulations were performed with a quasi deterministic radio channel generator (QuaDRiGa) using the defined parameters. The channel capacity, Demmel condition number, and relative condition numbers are used as the comparison metrics between reconstructed and modeled channels. The results show that the reconstructed MIMO channel matches the simulated MIMO channel well. / Tiivistelmä Tämä väitöskirja esittää parametroinnit kolmiulotteiselle geometriaan perustuvalle stokastiselle radiokanavamallille 10 GHz:n taajuusalueella perustuen mitattuun radiokanavaan. Väitöskirja koostuu kolmesta pääalueesta: radiokanavamittaukset, radiokanavamallin parametrien määrittäminen ja mallin validointi. Aluksi kuvataan kaksikerroksisessa aula ja kaupunkipiensolu ympäristöissä monilähetin monivastaanotin (MIMO) järjestelmällä tehdyt kanavamittaukset. Mittaukset tehtiin vektoripiirianalysaattorilla ja kaksoispolaroiduilla virtuaaliantenniryhmillä 500 MHz kaistanleveydellä. Mittausdata jälkikäsiteltiin käyttämällä ESPRIT-algoritmia ja jälkikäsitelty data varmennettiin osittain deterministisellä mittausympäristön karttaan pohjautuvalla radiokanavamallilla. Tulokset osoittivat hyvän yhteensopivuuden mitattujen ja mallinnettujen moniteiden välillä. Lisäksi toteuttiin yksi-lähetin yksi-vastaanotin mittaukset ulko-sisä etenemisympäristössä monikerroksisen lasin ja betoniseinän läpi. Tilastollinen analyysin avulla määritettiin täysi kolmiulotteinen kuvaus radioaallon etenemiskanavasta näköyhteys ja näköyhteydettömässä tilanteissa. Moniteiden suuremmista vaimennuksista johtuen viive ja kulmahajonnat ovat pienemmät verrattaessa matalempiin taajuuksiin. Peiliheijastus on diffuusisirontaa merkittävämpi radioaallon etenemismekanismi johtaen pienempiin klustereiden kulmahajeisiin matalempiin taajuuksiin verrattuna. Monikerroksisen lasin läpäisyvaimennus on samankaltainen kuin alemmilla taajuuksilla, kun sitä vastoin betoniseinän vaimennus on muutaman desibelin suurempi kuin alemmilla taajuuksilla. Lopulta geometriaan perustava stokastinen radiokanavamalli validoidaan määritellyillä parametreilla. MIMO kanava uudelleen rakennetaan lisäämällä kolmiulotteiset antennien säteilykuviot estimoituihin radioaallon etenemisteihin. Vastaavasti radiokanava simuloidaan näennäisesti deterministisellä radiokanavageneraattorilla (QuaDRiGa) käyttäen määriteltyjä mallin parametreja. Kanavakapasiteettia, Demmel ehtolukua ja suhteellista ehtolukua käytetään vertailumittareina uudelleen rakennetun ja simuloidun kanavan välillä. Tulosten perusteella uudelleen rakennettu MIMO kanava on yhteensopiva simuloidun radiokanavan kanssa.

Abstract In this thesis, novel data and channel estimation methods are proposed and analyzed for low-complexity short-range communication (SRC) radios. Low complexity is challenging to achieve especially in very wideband or millimeter-wave SRC radios where phase recovery and energy capture from numerous multipaths easily become a bottleneck for system design. A specific type of transceiver is selected using pulse amplitude modulation (PAM) at the transmitter and energy detection (ED) at the receiver, and it is thus called an ED-PAM system. Nonnegative PAM alphabets allow using an ED structure which enables a phase-unaware detection method for avoiding complicated phase recovery at the receiver. Moreover, the ED-PAM approach results in a simple multipath energy capture, and only one real decision variable, whose dimension is independent of the symbol alphabet size, is needed. In comparison with optimal phase-aware detection, the appealing simplicity of suboptimal ED-PAM systems is achieved at the cost of the need for a higher transmitted signal energy or shorter link distance for obtaining a sufficient signal-to-noise ratio (SNR) at the receiver, as ED-PAM systems are more vulnerable to the effects of noise and interference. On the other hand, the consequences of requiring a higher SNR may not be severe in the type of SRC scenarios where a sufficient received SNR is readily available due to a short link distance. Furthermore, significant interference can be avoided by signal design. However, what has slowed down the development of ED-PAM systems is that efficient symbol decision threshold estimation and related error probability analysis in multipath fading channels have remained as unsolved problems. Based on the above observations, this thesis contributes to the state-of-the-art of the design and analysis for ED-PAM systems as follows. Firstly, a closed-form near-optimal decision threshold selection method, which adapts to a time-varying channel gain and enables an arbitrary choice of the PAM alphabet size and an integer time-bandwidth product of the receiver filters, is proposed. Secondly, two blind estimation schemes of the parameters for the threshold estimation are introduced. Thirdly, analytical error probability evaluation in frequency-selective multipath fading channels is addressed. Special attention is given to lognormal fading channels, which are typically used to model very wideband SRC multipath channels. Finally, analytical error probability evaluation with nonideal parameter estimation is presented. The results can be used in designing low-complexity transceivers for very wideband and millimeter-wave wireless SRC devices of the future. / Tiivistelmä Tässä työssä esitetään ja analysoidaan uusia data- ja kanavaestimointimenetelmiä, joiden tavoitteena on yksinkertaistaa lähikommunikaatiota (short-range communication, SRC) langattomien laitteiden välillä. SRC-radioiden yksinkertainen toteutus on poikkeuksellisen haasteellista silloin, kun käytetään erittäin suurta kaistanleveyttä tai millimetriaaltoalueen tiedonsiirtoa. Tällöin vastaanottimen yksinkertaisen toteutuksen voivat estää esimerkiksi kantoaallon vaiheen estimointi ja signaalienergian kerääminen lukuisilta kanavan monitiekomponenteilta. Näistä lähtökohdista valitaan SRC-radion järjestelmämalliksi positiiviseen pulssiamplitudimodulaatioon (pulse amplitude modulation, PAM) perustuva lähetin ja energiailmaisimeen (energy detection, ED) perustuva vastaanotin. ED-PAM-järjestelmän ei tarvitse tietää vastaanotetun signaalin vaihetta ja signaalienergian kerääminen tapahtuu yksinkertaisen diversiteettiyhdistelytekniikan avulla. Lisäksi ilmaisuun tarvitaan vain yksi reaalinen päätösmuuttuja, jonka dimensio on riippumaton PAM-tasojen määrästä. ED-PAM-tekniikan yksinkertaisuutta optimaaliseen vaihetietoiseen ilmaisuun verrattuna ei saavuteta ilmaiseksi. Yhtenä rajoituksena on alioptimaalisen ED-PAM-tekniikan luontainen taipumus vahvistaa kohinan ja häiriöiden vaikutusta symbolin päätöksenteossa. Kohinan vahvistus ei välttämättä ole suuri ongelma niissä SRC-radioissa, joissa pienen linkkietäisyyden johdosta riittävä signaali-kohinasuhde vastaanottimessa voidaan kohinan vahvistuksesta huolimatta saavuttaa. Myös häiriöiden vahvistuksen vaikutusta voidaan tehokkaasti vähentää signaalisuunnittelulla. Joka tapauksessa ED-PAM-tekniikan käyttöönottoa on hidastanut tehokkaiden symbolipäätöskynnysten estimointi- ja analysointimenetelmien puuttuminen. Edellä mainitut havainnot ovat motivoineet löytämään uusia suunnittelu- ja analyysimenetelmiä ED-PAM-järjestelmille seuraavasti. Symbolipäätöskynnysten estimointiin johdetaan lähes optimaalinen suljetun muodon menetelmä, joka kykenee adaptoitumaan muuttuvassa kanavassa ja valitsemaan mielivaltaisen kokonaisluvun sekä PAM-tasojen määrälle että vastaanottimen aika-kaistanleveystulolle. Lisäksi esitetään kaksi sokeaa päätöskynnysten estimointimenetelmää, jotka eivät tarvitse redundanttista opetussignaalia. Työn toisessa osassa ED-PAM-järjestelmän symbolivirhesuhdetta analysoidaan taajuusselektiivisessä monitiekanavassa. Analyysissä keskitytään log-normaalijakauman mukaan häipyvään kanavaan. Seuraavaksi analyysia laajennetaan ottamalla mukaan epäideaalisten kynnysarvojen estimoinnin vaikutus. Saavutettuja tuloksia voidaan hyödyntää erittäin laajakaistaisten ja millimetriaaltoalueen SRC-laitteiden suunnittelussa.

Dans les systèmes de communications sans fil, le canal de transmission entre les antennes d’émission et de réception est l’une des principales sources de perturbation pour le signal. Les modulations multiporteuses, telles que l’OFDM (pour orthogonal frequency division multiplexing), sont très robustes contre l’effet des multi-trajets, et permet de retrouver le signal émis avec un faible taux d’erreur, quand elles sont combinées avec un codage canal. L’estimation de canal joue alors un rôle clé dans les performances des systèmes de communications. Dans cette thèse, on étudie des techniques fondées sur les estimateurs LS (pour least square, ou moindres carrés) et MMSE (pour minimum mean square error, ou erreur quadratique moyenne minimum). La technique MMSE est optimale, mais est beaucoup plus complexe que LS, et nécessite la connaissance a priori des moments de second ordre du canal et du bruit. Dans cette présentation, deux méthodes permettant d’atteindre des performances proches de LMMSE en évitant ses inconvénients sont étudiées. Une troisième partie étudie quant à elle les erreurs d’estimation dues aux interpolations. / In wireless communications systems, the transmission channel between the transmitter and the receiver antennas is one of the main sources of disruption for the signal. The multicarrier modulations, such as the orthogonal frequency division multiplexing (OFDM), are very robust against the multipath effect, and allow to recover the transmitted signal with a low error rate, when they are combined with a channel encoding. The channel estimation then plays a key role in the performance of the communications systems. In this PhD thesis, we study techniques based on least square (LS) and minimum mean square error (MMSE) estimators. The MMSE is optimal, but is much more complex than LS, and requires the a priori knowledge of the second order moment of the channel and the noise. In this presentation, two methods that allow to reach a performance close to the one of LMMSE while getting around its drawback are investigated. In another way, a third part of the presentation investigates the errors of estimation due to the interpolations.

No description available.