L'objectif de cette thèse est de traiter la problématique de la caractérisation et du traitement de la variabilité de l'exposition humaine aux ondes radio à travers l'utilisation de la dosimétrie numérique. En effet, si les progrès dans le domaine du calcul hautes performances ont contribué à significativement réduire les temps de simulation pour l'évaluation de l'exposition humaine, ce calcul du débit d'absorption spécifique reste un processus coûteux en temps. Avec la grande variabilité des usages, cette contrainte fait que la prise en compte de l'influence de paramètres d'entrée aléatoires sur l'exposition ne peut se faire par des méthodes classiques telles que les simulations de Monte Carlo. Nous proposons dans ces travaux deux approches pour répondre à cette problématique. La première s'appuie sur l'utilisation et l'hybridation de méthodes de construction de modèles de substitution afin d'étudier l'influence globale des paramètres d'entrée. La deuxième vise à l'évaluation efficace et parcimonieuse des quantiles à 95% des distributions de sortie et s'appuie sur le développement d'une méthode de planification d'expériences adaptative et orientée couplée à la construction de modèles de substitution. Les méthodes proposées dans ce manuscrit sont comparées et testées sur des exemples analytiques et ensuite appliquées à des problèmes concrets issus de la dosimétrie numérique / The purpose of this thesis is to deal with the problem of the management and the characterization of the variability of the human exposure to radio frequency waves through the use of the numerical dosimetry. As a matter of fact, if the recent advances in the high performance computing domain led to reduce significantly the simulation duration for the evaluation of the human exposure, this computation of the specific absorption rate remains a time-consuming process. With the variability of the usage, this constraint does not allow the analysis of the influence of random input parameters on the exposure to be achieved with classical approaches such as Monte Carlo simulations. In this work, two approaches are proposed to address this problem. The first one is based on the use and the hybridization of construction methods of surrogate models in order to study the global influence of the input parameters. The second one aims at assessing efficiently the 95th-percentiles of the output distributions in a parcimonous way. It is based on the development of an adaptive and oriented methodology of design of experiments combined with the construction of surrogate models. In this manuscript, the proposed methods are compared and tested on analytical examples and then applicated to full-scale problems from the numerical dosimetry

Les recherches en dosimétrie numérique non-ionisante ont connu des avancées considérables notamment grâce à la capacité de calcul croissante des ordinateurs. Ces dernières années, la prise en compte de la variabilité dans la dosimétrie est devenue un enjeu majeur. Les sources de variabilité sont nombreuses, parmi elles, les conditions d'exposition au rayonnement électromagnétique peuvent induire des doses absorbées très différentes. La modélisation de canal de propagation en environnement intérieur a permis d'avoir une connaissance précise des conditions d'exposition d'un corps humain plongé dans cet environnement. Cette thèse a pour but de développer une méthode statistique du calcul de la dose absorbée par le corps et d'adapter le modèle de canal stochastique à la dosimétrie. L'étude statistique de l'exposition a révélé la nécessité d'obtenir les valeurs de Débit d'Absorption Spécifique corps entier dans le cas d'expositions à une onde plane pour tous les angles d'incidence possibles. Compte tenu des temps de calcul particulièrement long en dosimétrie, une méthode d'interpolation efficace, le krigeage, des valeurs de Débit d'Absorption Spécifique a été mise en oeuvre. L'analyse de sensibilité aux paramètres du canal des moments du Débit d'Absorption Spécifique a permis de connaître l'impact de chacun de ces paramètres. Le modèle de canal a pu être simplifié et donc adapté à la dosimétrie tout en quantifiant l'erreur d'approximation qu'implique cette simplification. Cette thèse répond à l'enjeu de la prise en compte de la variabilité en dosimétrie dans un environnement complexe. / Research in non-ionizing numerical dosimetry has been improved thanks to high calculation capacity of computers. These years, integrating variability in the field of dosimetry has become a major issue. Sources of variability are numerous; among them, there are the exposure conditions to electromagnetic radiation which can lead to very different absorbed doses. Indoor channel modeling enables to have a deep knowledge of the exposure conditions of a human body located inside this indoor environment. The aim of this thesis is to develop a statistical method of calculation of the absorbed dose by the human body and to adapt the stochastic channel model to dosimetry. The statistical study of exposure reveals the need to obtain Specific Absorption Rate values for a plane wave exposure for all possible angles of incidence. Taking into account that computation in dosimetry is time consuming, an efficient interpolation method, kriging method, is implemented in order to get whole body Specific Absorption Rate values. Kriging method enables to obtain Specific Absorption Rate for all possible angles of incidence and then to calculate expectation and variance of Specific Absorption Rate. Sensitivity Analysis of expectation and variance to the statistical channel parameters reveals the impact of each parameter. The channel model has been simplified and then adapted to dosimetry by estimating the approximation error induced by this reduction. This thesis answers to the issue of integrating variability in dosimetry in a complex environment and develop the tools that open a new path in studying exposure in any complex environment.

Due to the great demand for high-speed wireless communication, Multi Input Multi Output (MIMO) antenna systems have been attracted rapidly increased attention. Therefore, the interaction between human body and MIMO antenna becomes an important issue, which will be studied in the present thesis. This thesis mainly focuses on the specific absorption rate (SAR) of the LTE MIMO antenna in mobile phone. It is different from the SISO antenna, and as the MIMO antenna has more operation modes and functions (MIMO, SISO, beam forming, etc.), more parameters need to be investigated for the SAR of the antenna. In this thesis, four designs of dual elements MIMO antenna (dual semi ground free PIFA, co-located antenna, dual OG PIFA in parallel position and dual OG PIFA in orthogonal position) are studied under four typical LTE frequency points (0.75GHz, 0.85GHz 1.9Ghz and 2.1/2.6GHz), with the effects of different chassis lengths (90mm, 110mm, 130mm and 150mm), the port’s states (terminated with 50ohm, open or short) and phase difference (0 degree to 360 degree). The SAR, when dual elements operate simultaneously, is also studied, which is evaluated by the SAR to PEAK Location Spacing Ratio (SPLSR). The simulations are run on both a SAM head phantom and a flat phantom by CST 2011, and measurements on flat phantom are carried out with iSAR and Dasy to verify the accuracy of our simulations. Besides SAR, the body loss of MIMO and SISO LTE antennas (MIMO: dual semi ground free PIFA and co-located Antenna; SISO: OG PIFA on top and OG PIFA at the bottom) in mobile phones are also studied in this thesis. The body loss and multiplex efficiency (MUX) with head and hand (H+H) in an increased cell phone case length (110mm, 130mm and 150mm) are investigated and discussed. Furthermore, some prominent problems in body loss study (body loss reduction, right and left hand problem, etc.) are also discussed in this thesis.

No description available.

No description available.

A grande maioria da população mundial está crescentemente exposta à radiação eletromagnética proveniente de fontes que muitas vezes estão localizadas nas proximidades do corpo. A radiação eletromagnética é considerada um agente possivelmente cancerígeno para as pessoas, classificação 2B indicada pela Organização Mundial da Saúde-OMS (WHO/IARC, 2011). Devido às preocupações em relação aos riscos associados a esta exposição existem normas que recomendam os valores máximos de exposição permitidos (ICNIRP, 1998; FCC, 2001). A correta avaliação das doses de radiação é, portanto, relevante. Este trabalho tem a finalidade de avaliar o impacto dos resultados do cálculo da dose da Taxa de Absorção Específica (SAR) em usuários expostos a radiação por tablets operando na faixa de radiocomunicações Wi-fi. Os três modelos existentes de cabeça humana utilizados serão um manequim homogêneo SAM phantom e dois modelos de cabeça realistas heterogêneos: um adulto masculino e uma criança masculina. Será também utilizado nas simulações um modelo masculino de criança que foi desenvolvido através de imagens de tomografia computadorizada (TC) pelo processo de segmentação feito no software AMIRA. Será utilizado um modelo genérico de tablet. Os parâmetros dosimétricos usados para simulação da SAR serão computados pelo software SEMCAD X que é baseado no Método das Diferenças Finitas no Domínio do Tempo (FDTD). Será criado também um código do Método FDTD através do software MATLAB que servirá para a escolha dos parâmetros do SEMCAD X. A distância entre o tablet e os modelos de cabeças varia de 50 mm a 300 mm. Os resultados da SAR serão comparados com os limites de exposição recomendados pelas normas internacionais. Também serão simuladas diferentes posições da antena no tablet. Da análise dos resultados foi constatado que os valores de SAR são muito baixos e todos os resultados ficaram dentro dos limites do psSAR recomendados pela FCC de 1,6 W/kg em cada 1 g de tecido e de 2 W/kg em cada 10 g de tecido estabelecidos pela ICNIRP. Comparando os valores de SAR do modelo SAM com o modelo DUKE, o modelo SAM se mostra conservador, porém quando a comparação é feita com as crianças o SAM deixa de ser conservador. / The vast majority of the world population is increasingly exposed to electromagnetic radiation from sources which are often located near to the body. Electromagnetic radiation is considered a possible carcinogen for people, classification 2B indicated by the World Health Organization-WHO (WHO/IARC, 2011). Due to concerns regarding the risks associated with this exposition there are regulations suggesting maximum allowed exposure values (ICNIRP, 1998; FCC, 2001). The correct evaluation of radiation doses is therefore relevant. This work aims to assess the impact of the results of the calculation of Specific Absorption Rate dose (SAR) in users exposed to radiation from tablets operating in the Wi-fi band. The three existing models of human head used are a homogeneous dummy SAM phantom and two heterogeneous realistic head models: a male adult and a male child. It will also be used in the simulations a male child model which was developed from computed tomography (CT) imaging using the AMIRA software for the segmentation process. A generic model of tablet is used. Dosimetric parameters used for simulation of the SAR are computed using the SEMCAD X software which is based on the Finite Difference Method in Time Domain (FDTD). A FDTD code was developed using the MATLAB software in order to help to choose the input SEMCAD X parameters. The distances between the tablet and the head of the models varies from 50 mm to 300 mm. SAR results are compared with the exposure limits recommended by international standards. Different antenna positions on the tablet are simulated too. Examining the results it was found that the SAR values are very low and all results are within the psSAR limits recommended by FCC (1,6 W/kg averaged over 1 g of tissue) and by ICNIRP (2 W/kg in 10 g of tissue). Comparing the SAR in the SAM model with the SAR in the DUKE model, the SAM model shows to be conservative. However, when compared with the children, the SAM is not conservative.

Dans cette thèse, nous proposons des modifications de conception d’antenne PIFA concernant la téléphonie mobile afin de réduire le taux de pénétration des ondes électromagnétiques dans la tête. Trois antennes ont été considérées: une antenne PIFA à bande unique fonctionnant à 1,8 GHz, une antenne PIFA à bande double fonctionnant à 900 MHz et à 1,8 GHz, et une antenne PIFA tri-bande à 900 MHz, 1800 MHz et 2400 MHz. La densité de courant de surface le long du plan de masse a été réduite en diminuant le rayonnement du champ électrique émis par l’antenne émettrice vers le corps, ce qui a entraîné une réduction de la densité de courant le long de la tête humaine ainsi que du SAR. Cette réduction du SAR a été réalisée en insérant plusieurs traitements de bords tels qu'une forme en U à chaque coin du plan de masse de l'antenne. Nous avons étudié trois types des antennes : la première, lorsque les bords U sont constitués de PEC, nous obtenons une réduction du SAR de 78,6%. La seconde utilise un type de liquide considéré comme absorbant. Avec ce liquide, deux antennes ont été analysées: une PIFA à bande unique fonctionnant à 1,8 GHz et une PIFA à bande double fonctionnant à 900 MHz et à1,8 GHz. Dans ce scénario, le bord en U est constitué d’eau, et nous avons constaté que la valeur maximale du SAR moyen sur 10 g de tissu était réduite à 73,52% pour une PIFA à bande unique, tandis que le pic de SAR a été réduit à 88,78% et 90,31% en cas de double bande. Afin de préserver la hauteur de l'antenne et de la rendre plus réaliste, un autre type d’antenne a été modélisé en insérant des arêtes en forme de U à chaque extrémité du plan de masse. Dans ce cas, la réduction du SAR a été la suivante: 74,5% à 1800 MHz pour une bande unique et environ 64,37%, 79,92% et 54,71% à 900 MHz, 1800 MHz et 2400 MHz pour une antenne tri-bande, et pour une moyenne de masse de tissu égale à 10g. Dans une dernière partie, cette thèse propose une antenne optimisée pour des transmissions de la station de base vers l'utilisateur d’un téléphone mobile, en exploitant la bande GSM-900. L’objectif principal de cette étude est de fournir une intensité de champ constante sur la surface de couverture, afin de minimiser certains effets biologiques considérés comme dangereux pour la santé humaine. / In this thesis, we propose new types of mobile handset PIFA antenna design to reduce the specific absorption rate (SAR) in human head. Three antennas were considered: a single band PIFA operating at 1.8 GHz, a dual band PIFA operating at 900 MHz and 1.8 GHz, and a Triband PIFA operating at 900MHz, GSM 1800MHz, and 2400MHz. Surface current density along the ground plane was decreased by reducing the near electric field radiation coming from the radiating antenna to the human head, which in turns leads to a reduction in current density along human head as well as the SAR. This SAR reduction is accomplished by inserting several edges treatment such as U-shape at each corner of the ground plane of antenna. In this scenario, we studied three types: first one, when U-edges are made of PEC, we get SAR reduction about 78.6%. Second one is a liquid type that is considered as more advanced type. For this type, two antennas are considered: a single band PIFA operating at 1.8 GHz, and a dual band PIFA operating at 900 MHz and 1.8 GHz. In this scenario, two cases were studied: first one when the U-edge is made of water, we noticed that the peak SAR value averaged over 10 g mass tissue was reduced to 73.52% in the case of single band PIFA, whereas the peak SAR was reduced to 88.78% and 90.31% in case of dual band. Second case, when U-edge filled with water, this enhanced structure was presented to overcome the manufacturing limitations of the first case design. Simulation results showed a better performance in terms of reflection coefficient and radiation pattern for both single-band and dual-band antennas. The results of SAR were also similar to the first case structure design for the single-band antenna. However, the SAR was slightly higher for the dual-band antenna, but it still remains very low (1.6 W/Kg at 900 MHz and 0.6 W/Kg at 1800 MHz). After that, for preserving the height of antenna, and making it more realistic to be applied, another type was done by inserting U-edges above each corner of the ground plane. In this case, we get SAR reduction as follows: 74.5% at GSM1800MHz in case of single band, and about 64.37%, 79.92%, and 54.71% at GSM900MHz, GSM1800MHz, and 2400 MHz in case of Tri-band averaged over 10g mass tissue. On the other hand, this thesis presents a modern and unprecedented safety antenna that transmits from the base station to the mobile user being operated at the GSM-900 band. The main objective of this study is to provide a constant field strength or a uniform coverage over a certain surface area, for preventing human health from several hazardous biological effects, and this strongly depends on the studied site (e.g. if it is densely urban, rural, open area, etc.). Our proposed design is synthesized using S.A.R.A (Synthesis of Array of Antenna) software. Besides, Atoll GSM planning tool is used to ensure the signal coverage and power strength for the safety antenna, and CST (Computer Simulation Technology) software for simulating the proposed antenna.

In this thesis, an internal mobile phone antenna with a distributed matching circuit for eight-band LTE/WWAN operation is presented. By selecting proper dimensions of the distributed matching circuit, the bandwidth of the proposed antenna can be greatly enhanced. Good radiation characteristics are also obtained, and the proposed antenna occupies only 60 ¡Ñ 10 ¡Ñ 3 mm^3 and is very suitable for slim mobile phone applications. Effects of the human body including user¡¦s head and hand are also studied, and the SAR and HAC issues are simulated and analyzed in this thesis.

A tablet computer antenna array for WWAN/LTE and LTE MIMO operations is proposed. The antenna array comprises a main antenna and an auxiliary antenna. The main antenna is an eight-band coupled-fed antenna, which can cover the GSM850/900/1800/1900/UMTS and LTE700/2300/2700 operations. The auxiliary antenna is a three-band antenna, which can cover the LTE700/2300/2500 operation and occupies a small size. Acceptable isolation between the main and auxiliary antennas has also been obtained, which makes it promising for the main and auxiliary antennas to perform LTE MIMO operation in the LTE700/2300/2500 bands. Effects of the internal tablet computer antenna on the user¡¦s body are also studied. The obtained results show that the antenna can meet the 1-g body SAR specification of less than 1.6 W/kg by selecting a proper distance between the antenna and the flat phantom. Also, since the user¡¦s body is a lossy material, a decrease in the antenna¡¦s radiation efficiency is observed when the user¡¦s body is in the vicinity of the internal antenna.

In this thesis, a comparison of the body SAR value of five planar WLAN antennas including two coupled-fed shorted monopole antennas, a coupled-fed loop antenna, a monopole antenna, and a monopole slot antenna for tablet PC applications are presented. A required minimum distance between the antenna and the testing flat phantom to meet the 1-g body SAR requirement of 1.6 W/kg has been determined. Results show that the body SAR results of the coupled-fed loop antenna are lowest among the five tested antennas. Detailed results and discussion are presented in this thesis. Also, a comparison of the body SAR value of a planar antenna and an antenna with 3-D bent structure are presented.