Systems Performance Engineering and Applied Queueing

Computer and telecommunication networks are large-scale systems consisting of many components that interact to provide a wide variety of services. The performance of these systems is dependent on how well various protocols coordinate and manage complex interactions in order to provide an acceptable quality of service. Systems performance engineering and applied queueing research involves the performance modeling of these complex systems in order to gain better understanding and insight into their behavior. It is motivated by a number of interesting performance problems, where solutions frequently require innovative and novel applications of queueing models and large scale simulation. Some specific research areas include: predictive models for resource demand in wireless networks, resource allocation and optimal routing in wireless mesh networks, and the development of efficient solution methodologies for queueing models of systems incorporating correlation over a finite time horizon. Various funded projects are described below.

Recent and Current Research

Scalable Performance Models for Large Scale Networks with Correlated Traffic

Principal Investigator(s): Ken Mitchell, Ph.D.
Co-Principal Investigator(s): Appie van de Liefvoort, Ph.D., Victor Wallace, Ph.D.
Funded by: The National Science Foundation (NSF)
Award number: ANI-0106640
Goal: Build powerful, efficient, and scalable models that capture network behavior for varying levels of abstraction and varying finite time scales while incorporating (approximations to) long-range dependence and heavy tails. This entails the development of network models that can be employed at several layers of abstraction, which will allow for a full understanding of network behavior, and will use (nearly-) decomposable techniques for state reduction and time scale resolution. These models will not solely rely on steady state results, but rather on a finite horizon that would allow for online monitoring and early detection of rare events.

A Study of TCP Performance over Mobile Ad Hoc Networks

Principal Investigator(s): Ken Mitchell, Ph.D.
Funded by: University of Missouri Research Board
Goal: Investigate transport layer behavior using both simulation and analytic studies that will accurately predict performance measures and performance bounds. We will identify and quantify the impact that newly proposed protocols designed for ad hoc networks have on TCP performance, and based on the knowledge gained from these studies, either develop enhancements to the TCP/IP suite or propose new transport layer protocols.

Network Modeling and Analysis: A Rational Series Approach

Principal Investigator(s): Appie van de Liefvoort, Ph.D.
Partner(s): Khosrow Sohraby, Ph.D.
Funded by: University of Missouri Research Board
Goal: Accurate modeling and prediction of network and system performance continuous to challenge network designers and operators, who can no longer rely on Markovian Models and Poisson characterizations. This research attempts to incorporate serial and local dependencies, as well as complex traffic patterns in both analytic and simulation models. Model identification and parameter estimation will be addressed as well to keep the dimensionality of such models minimal.