Datenbestand vom 13. Juni 2019

Warenkorb Datenschutzhinweis Dissertationsdruck Dissertationsverlag Institutsreihen     Preisrechner

WICHTIGER HINWEIS
DER VERLAG IST IN DER ZEIT VOM 12.06.2019 BIS 23.06.2019 AUSCHLIESSLICH PER EMAIL ERREICHBAR.

aktualisiert am 13. Juni 2019

ISBN 9783843914727

Euro 84,00 inkl. 7% MwSt


978-3-8439-1472-7, Reihe Informatik

Christopher Mutschler
Latency Minimization of Order-Preserving Distributed Event-Based Systems

229 Seiten, Dissertation Universität Erlangen-Nürnberg (2014), Softcover, A5

Zusammenfassung / Abstract

Nowadays sensors are increasingly deployed in many kinds of applications and deliver streams of data that they continuously collect. This data is significant as it provides important information in real time. However, without a fast processing of this information the sensors only provide a pointless stream of data. Hence, there is a need for automatic processing to extract meaningful information within an acceptable amount of time. This thesis describes techniques that allow a fast analysis of streaming data. Real-time Locating Systems (RTLSs) or Radio Frequency Identification (RFID) systems provide several thousand position events per second. Event-based systems (EBSs) meet the high performance requirements and are a powerful technique for a reactive analysis of such data streams. Detection algorithms are divided up into several comparatively small event detectors (EDs), become inherently scalable through distribution, and are easy to maintain because of their reduced software complexity. Such event detectors communicate by messages, i.e., events, over an event processing middleware and are hierarchically linked to detect the final events of interest.

Since partial results, i.e., events, are generated at different points in the system they are no longer timely synchronized. However, the algorithms implemented in the event detectors assume a timely ordered event stream as they try to detect interaction patterns. It is never a viable solution to process events out of order. This puts a significant workload to the underlying middleware. A-priori estimations of reordering parameters cannot include runtime information about object and system behavior, and thus the event loads, and must hence be set too conservatively in order to avoid system failures caused by ordering mistakes. But this often results in high detection latencies.

This thesis describes how to optimally adapt to variations in the observed environment to minimize detection delays at runtime. We show how out-of-order events are transparently reordered with low latency at each node so that event detectors may process them in a correct order. A speculative processing exploits unused system resources and reduces detection latency to a minimum. We further present a technique to migrate event detectors between nodes at runtime and show how to optimize their detection latency introduced by networking delays in a distributed system environment.