Bitte benutzen Sie diese Kennung, um auf die Ressource zu verweisen: http://dx.doi.org/10.18419/opus-9615
Autor(en): Roy Chowdhury, Sujata
Titel: Packet scheduling algorithms for a software-defined manufacturing environment
Erscheinungsdatum: 2015
Dokumentart: Abschlussarbeit (Master)
Seiten: ix, 69, iii
URI: http://nbn-resolving.de/urn:nbn:de:bsz:93-opus-ds-96326
http://elib.uni-stuttgart.de/handle/11682/9632
http://dx.doi.org/10.18419/opus-9615
Zusammenfassung: With the vision of Industry 4.0, Internet of things (IoT) and Internet of Services (IoS) are making their way to the modern manufacturing systems and industrial automation. As a consequence, modern day manufacturing systems need wider product variation and customization to meet the customer's demands and survive in the competitive markets. Traditional, dedicated systems like assembly lines cannot adapt the rapidly changing requirement of today's manufacturing industries. A flexible and highly scalable infrastructure is needed to support such systems. However, most of the applications in manufacturing systems require strict QoS guarantees. For instance, time-sensitive networks like in industrial automation and smart factories need hard real-time guarantees. Deterministic networks with bounded delay and jitter are essential requirement for such systems. To support such systems, non-deterministic queueing delay has to be eliminated from the network. To this end, we present Time-Sensitive Software-Defined Networks (TSSDN) with a logically centralized controller which computes transmission schedules based on the global view of the network. SDN control logic computes optimized transmission schedules for the end hosts to avoid in network queueing delay. To compute transmission schedules, we present Integer Linear Programming and Routing and Scheduling Algorithms with heuristics that schedule and route unicast and multicast flows. Our evaluations show that it is possible to compute near optimal transmission schedules for TSSDN and bound network delays and jitter.
Enthalten in den Sammlungen:05 Fakultät Informatik, Elektrotechnik und Informationstechnik

Dateien zu dieser Ressource:
Datei Beschreibung GrößeFormat 
Thesis_Final 1.0.pdf1,1 MBAdobe PDFÖffnen/Anzeigen


Alle Ressourcen in diesem Repositorium sind urheberrechtlich geschützt.