Photo by ©FAU/Erich Malter

OPEN FAU

Online publication system of Friedrich-Alexander-Universität Erlangen-Nürnberg

The online publication system OPEN FAU is the central publication platform for Open Access publishing for all members of Friedrich-Alexander-Universität. Qualified works from research and teaching may be published here free of charge, either as a primary or secondary publication. The full texts are permanently available worldwide and are findable and citable via catalogues and search engines.


To search for documents in OPEN FAU, please select "Search" (via the magnifying glass at the top right); this will provide you with various search options. If you want to publish a document, go to "Login" and "My Publications". Then drag you document into the field provided and enter the metadata. In just a few steps, you can submit your document. Please note our guidelines, the publication contract and FAQs.

 

Recent Submissions

Article
Open Access
Primärprozesse der Photosynthese - kinetische Modellierung
(Humboldt-Universität zu Berlin, 1989-05-01) Voigt, Jochen; Maciy, Katina; Pöschel, Thorsten; Schimansky-Geier, Lutz
Article
Open Access
A Multiprocessor System for High Speed Simulation of Neural Networks
(Akademie-Verlag, 1990) Boseniuk, Thorsten; van der Meer, Marius; Pöschel, Thorsten
A special purpose low-cost multiprocessor system for the high-speed simulation of error back-propagation (and related) algorithms will be proposed. It will be shown that the processing time rises with 0(LN) (L-number of layers in the network, N-number of elements per layer), while universal serial machines need 0(LN^2). A detailed hardware description will be given.
Article
Open Access
Modellierung Neuronaler Netzwerke
(Humboldt-Universität zu Berlin, 1989-05-01) Boseniuk, Thorsten; Pöschel, Thorsten
Doctoral thesis
Open Access
A Unified Approach to RSS-Based Localization in Wireless Sensor Networks
(2024) Nowak, Thorsten; Thielecke, Jörn
Nowadays, many applications demand precise location information. One of those applications is wildlife monitoring, e.g., tracking bats in a forest. Radio-based localization is very popular in the context of wildlife monitoring. However, radio-based localization suffers from multipath propagation that arises in real-world scenarios due to reflections and scattering. Three major aspects in radio-based localization are targeted in the thesis. These are performance metrics, multipath robustness and adaptivity. Users of positioning systems are interested in the expected performance of the localization. Thus, performance metrics are one of the major issues in the design of positioning systems. Estimation theory addresses the problem of performance in radio-based localization. However, classical estimation theory has some weaknesses when considering received signal strength (RSS)-based direction finding. Therefore, information-theoretic measures are derived to overcome these weak points. Information-theoretic performance metrics allow for characterization and optimization of ambiguous localization sensors in RSS-based direction finding. Furthermore, they allow for a unified design approach considering the individual sensors, the network topology as well as the impact of multipath propagation. Multipath propagation is still a major issue in radio-based localization. To provide multipath robustness two mitigation schemes applicable to RSS-based direction finding are presented. One is a probabilistic approach based on knowledge of channel parameters, e.g., the angular spread. Statistical knowledge is integrated into the process of RSS-based direction finding such that direction estimates are unbiased in a multipath scenario. However, the probabilistic approach has some drawbacks. On the one hand it requires prior knowledge of channel parameters which is rarely available in real-life scenarios. On the other hand probabilistic mitigation fails in severe multipath environments. Hence, besides probabilistic mitigation, a frequency-domain multipath resolution approach is presented. Applying super-resolution techniques unbiased direction estimates are achieved in severe multipath environments. In contrast to common broadband direction estimation, the presented approach does not require phase-coherent receive channels. Hence, the proposed superresolution approach is applicable to low-power networks and brings accurate positioning to small-sized and energy-efficient sensor nodes. Adaptivity addresses the need of tuning the localization system parameters to the current environment. Especially when considering the tracking of highly agile animals, like bats, positioning systems have to adapt to different surroundings very quickly. The statistical channel properties, which form the basis for probabilistic multipath mitigation, change within seconds. Therefore, tracking filters need to estimate position and channel parameters simultaneously. Within this thesis adaptive filters are presented that allow for simultaneous state and channel parameter estimation. Thus, there is no need to parameterize the tracking filters which enables the use in unknown changing environments. The three aspects performance metrics, multipath robustness and adaptivity enhance the applicability of radio-based localization systems in real-world applications. Predictable performance, multipath robustness, and adaptivity are a step towards plug & play localization systems which enables the usage of tracking systems in daily work of biologists to understand complex biological processes such as migration, foraging, and social interactions.
Article
Open Access
Computer simulieren verstopfte Rohre
(Forschungszentrum Jülich, 1993-03-01) Pöschel, Thorsten; Puhl, Harald
Computersimulationen gewinnen in Forschung und Industrie immer mehr an Bedeutung; sie ersetzen und ergänzen experimentelle Untersuchungen oder erleichtern das Verständnis von Ergebnissen. Solche Simulationen werden bekanntlich auch im Höchstleistungsrechenzentrum (HLRZ) durchgeführt. Mitarbeiter des HLRZ haben in der Vergangenheit unter anderem den Fluß von granularen Medien durch Rohre berechnet. Dabei konnte die Frage beantwortet werden, warum z. B. beim Ausfluß von Sand oder Körnern aus Silos Verstopfungen auftreten. Der Grund sind unregelmäßige Dichtewellen; ihre Berechnung kann zur Lösung zahlreicher industrieller Probleme beitragen.