Algorithms and Metrics for New Generation Data Stream Management Systems

University Of Pittsburgh, Pittsburgh PA

Investigators

Abstract

IIS-0534531 Panos K. Chrysanthis <panos@cs.pitt.edu> University of Pittsburgh Algorithms and Metrics for New Generation Data Stream Management Systems The goal of this project is to design a new generation of data stream management systems (DSMSs), with equal emphasis on optimizing performance and enhancing functionality. New generation DSMSs simplify the development of a wide range of monitoring applications, with diverse requirements. Monitoring applications are core components in scientific exploration, health alerting, environmental monitoring, and business support systems. This project reexamines all four critical components of a DSMS: query scheduler, load shedder, query processor, and data dissemination modules. The two key innovations of this project are: (1) it looks at how these four modules can be integrated to work in combination, instead of making in isolation decisions that have a significant impact on the overall performance; and (2) this project formalizes QoS/QoD metrics for DSMSs and develops algorithms designed to optimize these metrics. In addition, the project plans include the analytical and experimental evaluation of the proposed algorithms and also the implementation and evaluation of a prototype system. This project provides opportunities for both graduate and undergraduate students to participate in the development of cutting edge technology. Results of this research, including software, data, and publications, will be disseminated via the project Web site at http://db.cs.pitt.edu/streams