Speaker: Khaled Harfoush , Computer Science, Boston University, Mass.
A Framework and Toolkit for the Effective Measurement and Representation of Internet Internal Characteristics
Abstract: The development and deployment of distributed network-aware applications and services require the ability to compile and maintain a model of the underlying network resources with respect to (one or more) metrics of interest. To be manageable, such models must be compact, and must enable a representation of properties along temporal, spatial, and measurement resolution dimensions. In the first part of my talk, I will propose MINT--a general analytical framework for the construction of such metric-induced models using end-to-end measurements. I will instantiate this framework for a number of metrics, including hop count, packet loss rate, and bottleneck bandwidth.
In the second part of my talk, I will focus on the instantiation of MINT for the bottleneck bandwidth metric. The extensive body of prior work on the measurement of bottleneck bandwidth has focused on two approaches measuring it hop-by-hop, and measuring it end-to-end along a path. Unfortunately, best-practice techniques for the former are inefficient and techniques for the latter are only able to observe bottlenecks visible at end-to-end scope. I will present new end-to-end probing methods which efficiently measure bottleneck bandwidth along arbitrary, targeted subpaths of a path in the network, including subpaths shared by a set of flows.
All of the above techniques and constructions have been tested using simulation, implementation, and Internet deployment experiments. Throughout my talk, I will present results from these experiments. Also, I will describe a number of practical applications which stand to benefit from the proposed techniques, especially in emerging, flexible network architectures such as overlay networks, ad-hoc networks, peer-to-peer architectures and massively accessed content servers.
Short Bio: Khaled Harfoush received his B.S. degree and M.S. degree in Computer Science from the Alexandria University in Alexandria, Egypt in 1992 and 1995 respectively. Currently, he is enrolled in the Ph.D. program in the Department of Computer Science at Boston University. The title of his dissertation is: "A Framework For End-to-End Characterization of Metric-Induced Network Topologies." His main research interest is to develop end-to-end network diagnosis techniques to uncover dynamic network properties (e.g. congestion information, bottleneck equivalence, loss rates, topology, etc.) in a near real-time fashion.
Host: H. Perros and G. Rouskas, Computer Science, NCSU
Colloquia Home Page.