Duisburg, Universiẗat, Habil.-Schr., 2000.

Stuttgart, Universiẗat, Diss., 2003. / Dateien im PDF-Format.

Erlangen, Nürnberg, Universiẗat, Diss., 2003.

Tübingen, Universiẗat, Diss., 1999. / Erscheinungsjahr an der Haupttitelstelle: 1999.

Münster (Westfalen), Universiẗat, Diss., 2001.

Hannover, Universiẗat, Diss., 2002.

archives@tulane.edu / 1 / Jack Green

La représentation par complexes simpliciaux fournit une description mathématique de la topologie d’un réseau de capteurs, c’est-à-dire sa connectivité et sa couverture. Dans ces réseaux, les capteurs sont déployés aléatoirement en grand nombre afin d’assurer une connectivité et une couverture parfaite. Nous proposons un algorithme qui permet de déterminer quels capteurs mettre en veille, sans modification de topologie, afin de réduire la consommation d’énergie. Notre algorithme de réduction peut être appliqué à tous les types de complexes simpliciaux, et atteint un résultat optimal. Pour les complexes simpliciaux aléatoires géométriques, nous obtenons des bornes pour le nombre de sommets retirés, et trouvons des propriétés mathématiques pour le complexe simplicial obtenu. En cherchant la complexité de notre algorithme, nous sommes réduits à calculer le comportement asymptotique de la taille de la plus grande clique dans un graphe géométrique aléatoire. Nous donnons le comportement presque sûr de la taille de la plus grande clique pour les trois régimes de percolation du graphe géométrique. Dans la deuxième partie, nous appliquons la représentation par complexes simpliciaux aux réseaux cellulaires, et améliorons notre algorithme de réduction pour répondre à de nouvelles demandes. Tout d’abord, nous donnons un algorithme pour la planification automatique de fréquences, pour la configuration automatique des réseaux cellulaires de la nouvelle génération bénéficiant de la technologie SON. Puis, nous proposons un algorithme d’économie d’énergie pour l’optimisation des réseaux sans fil. Enfin, nous présentons un algorithme pour le rétablissement des réseaux sans fil endommagés après une catastrophe. Dans ce dernier chapitre, nous introduisons la simulation des processus ponctuelsdéterminantaux dans les réseaux sans fil. / Simplicial complex representation gives a mathematical description of the topology of a wireless sensor network, i.e., its connectivity and coverage. In these networks, sensors are randomly deployed in bulk in order to ensure perfect connectivity and coverage. We propose an algorithm to discover which sensors are to be switched off, without modification of the topology, in order to reduce energy consumption. Our reduction algorithm can be applied to any type of simplicial complex and reaches an optimum solution. For random geometric simplicial complexes, we find boundaries for the number of removed vertices, as well as mathematical properties for the resulting simplicial complex. The complexity of our reduction algorithm boils down to the computation of the asymptotical behavior of the clique number of a random geometric graph. We provide almost sure asymptotical behavior for the clique number in all three percolation regimes of the geometric graph. In the second part, we apply the simplicial complex representation to cellular networks and improve our reduction algorithm to fit new purposes. First, we provide a frequency auto-planning algorithm for self-configuration of SON in future cellular networks. Then, we propose an energy conservation fot the self-optimization of wireless networks. Finally, we present a disaster recovery algorithm for any type of damaged wireless network. In this last chapter, we also introduce the simulation of determinantal point processes in wireless networks.Simplicial complex representation gives a mathematical description of the topology of a wireless sensor network, i.e., its connectivity and coverage. In these networks, sensors are randomly deployed in bulk in order to ensure perfect connectivity and coverage. We propose an algorithm to discover which sensors are to be switched off, without modification of the topology, in order to reduce energy consumption. Our reduction algorithm can be applied to any type of simplicial complex and reaches an optimum solution. For random geometric simplicial complexes, we find boundaries for the number of removed vertices, as well as mathematical properties for the resulting simplicial complex. The complexity of our reduction algorithm boils down to the computation of the asymptotical behavior of the clique number of a random geometric graph. We provide almost sure asymptotical behavior for the clique number in all three percolation regimes of the geometric graph. In the second part, we apply the simplicial complex representation to cellular networks and improve our reduction algorithm to fit new purposes. First, we provide a frequency auto-planning algorithm for self-configuration of SON in future cellular networks. Then, we propose an energy conservation fot the self-optimization of wireless networks. Finally, we present a disaster recovery algorithm for any type of damaged wireless network. In this last chapter, we also introduce the simulation of determinantal point processes in wireless networks.

Note:

Continuous monitoring of physiological parameters such as pH and oxygen (O2) are of great importance in determining the health status of a patient. Arterial blood gas analysis is a current clinical method used to measure pH, PCO2, PO2, and the concentration of variety of ions, typically with blood withdrawn from an artery. The need for robust, and a rapidly responding technology to enable bed-side monitoring has driven considerable efforts to produce better sensor devices. Optical sensing systems have experienced rapid growth, with drivers including low-cost optical fibres, and the availability of miniature optical set-ups (light sources, detectors, etc.). Herein, polymer-based optical fibre sensors for pH and O2 sensing were developed. The pH and/or oxygen reporters were immobilised at the end of an optical fibre by photo-polymerisation, and their performance in measuring pH and O2 concentration investigated. pH sensing was based on fluorescence detection using single excitation/single emission (Chapter 2), and single excitation/dual emission (Chapter 3). O2 sensing was based on the luminescence quenching of metalloporphyrins by oxygen (Chapter 4). In the last chapter, the in vivo applicability of an O2 sensor was investigated by measuring O2 level changes inside an ex vivo lung.