Performance Modeling and Analysis of Real-Time Distributed Systems

University Of Illinois At Chicago, Chicago IL

Investigators

Abstract

The objective of this research is to investigate formal modeling and analysis methods that are applicable to large-scale real-time distributed systems, such as networked virtual environments. The research focuses on the development of techniques and tools for performance analysis of network protocols and Quality-of-Service requirements analysis of the network connecting distributed real-time applications. It is proposed to work on the following tasks: 1) Develop (extended fuzzy-timing Petri net) models for protocols in each layer of network architecture; 2) Conduct network performance analysis and evaluate network effects on several distributed real-time applications; 3) Focus on the development of new techniques to deal with the state explosion problem; and 4) Maintain an open library of models for real-time distributed applications. The significance and impact of the research are: 1) Provide an integrated framework to conduct performance analysis for large-scale real-time distributed systems with high efficiency; 2) Help designers of distributed real-time systems in testing, planning, and improving their work; and 3) Help students learn formal modeling and analysis methods for network-based software design via graphic interface and simulation.