Design and analysis of congestion control for high-speed networks.

Abstract

In this thesis, we first use the fluid flow modeling technique to build an end-to-end model and a hop-by-hop model for the high-speed networks respectively. We then use the control theory to design distributed rate-based traffic controllers to flow-regulate the best-effort service (e.g., ABR) traffic and guaranteed service traffic through a high-speed switch. The controllers are distributed among the source nodes and have a very simple structure. Their local controller at each source node is open-loop stable and only requires the knowledge of the queue length at the bottleneck switch. We say that the network is open-loop stable if the feedback loop breaks. Based on the simulation results obtained from OPNET and MATLAB, we finally show that the controllers are fair and can stabilize the network with long variable delays. Our controllers overcome the oscillation problem which exists in most of current existing rate-based controllers.