Thèse de doctorat
Résumé : The smart environments and the connected human seems to be the future of wireless communications. The development of new frequency bands in the millimeter range will allow us to create high data rate communications which will led to the Wireless Body Environment Networks. In this kind of scenarios, it is expected that the user and the environment will interact. In order to develop such new applications, it is necessary to first study the propagation mechanisms and then, the communication channel underlying body centric environments. This thesis treats of channel models for 60 GHz Body Area Networks and more particularly of three kinds of scenarios: (i) the communication between an external base station and a worn node (off-body); (ii) the communication between two worn nodes (on-body); the communication between an external base station and a hand-held device (near-body). An indoor off-body channel model is numerically proposed and implemented. The model is based on the IEEE 802.11ad indoor standard channel at 60 GHz and a fast computation solution of the scattering of a plane wave by a circular cylinder. The model is developed for two orthogonal polarizations and the communications performances are studied. The on-body propagation is studied for two different configurations: line-of-sight and non-line-of-sight communications on the body. These scenarios led to different solutions for the channel knowing as, respectively, Norton’s equations and creeping formulations. These solutions are obtained using simplified geometries which has been experimentally validated. Further, in order to improve the propagation on the human body, a technique using metallic plates has been proposed. This technique has been theoretically studied using Milligton’s equations and experimentally assessed on a flat phantom with the properties of the human skin. The proposed method allows to save up to 20 dB. Finally, the near-body communication scenario has been introduced and studied. The near-body region is extended from 5 to 30 cm away of the user body which corresponds to the arm’s reach and models a handheld device. A numerical algorithm has been proposed to model indoor near-body environments. Also, a special has been given to statistical body shadowing. It has been shown that the fading follows a Two-Wave Diffuse Power distribution.