Rethinking Internet Traffic Management: From Multiple Decompositions to a Practical Protocol
Report ID: TR-774-07Author: Bresler, Ma'ayan / Chiang, Mung / Rexford, Jennifer / He, Jiayue
Date: 2007-02-00
Pages: 14
Download Formats: |PDF|
Abstract:
In the Internet today, traffic management spans congestion control (at end hosts), routing protocols (on routers) and traffic engineering (by network operators). Historically, this division of functionality slowly evolved without a conscious design. In this paper, we perform a top-down redesign of traffic management using recent innovations in optimization theory. First, we propose an objective function that captures the goals of end users and network operators alike. Using all known optimization decomposition techniques, we generate four distributed algorithms where sources adapt their sending rates along multiple paths, based on different kinds of feedback from the links. Optimization theory guarantees that these algorithms converge to a stable and optimal point, and simulations allow us to compare rate of convergence, robustness to tunable parameters, and performance under realistic traffic. Combining the best features of the algorithms, we construct TRUMP: a traffic management protocol that is distributed, adaptive, robust, flexible and easy to manage. We show that using optimization decompositions as a foundation, simulations as a building block, and human intuition as a guide can be a principled approach to protocol