Delayed congestion response protocols

Due to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to digital@library.tamu.edu, referencing the URI of the item.Includes bibliographical references (leaves 61-63).Issued also on microfiche from Lange Micrographics.There has been a surge of new real-time multimedia applications over the IP network which cannot use TCP because the congestion control algorithm of TCP results in drastic variations in the sending rate, which in turn affect the user perceived quality. As a result several of these applications use UDP. The fact that UDP does not respond to congestion in any way has raised several questions about possible congestion collapse of the Internet and fairness to existing TCP-based applications. Several congestion control algorithms suited for such applications have been proposed which claim bandwidth in a "TCP-friendly" manner. However all of them reduce their sending rate once and as soon as allowed by the protocol design, when a packet drop occurs. In this thesis we have proposed a novel scheme by which response to congestion is deliberately delayed by [] RTTs. We have provided the framework for this class of protocols in general and examined three cases in particular. For these three cases we have developed analytical models and derived the conditions under which they are TCP-friendly. With these conditions we have run simulations on the ns-2 platform to show that they are indeed TCP-friendly. By showing that delayed congestion response is possible, we have laid the ground work for the development of a whole new class of protocols which are not just TCP-friendly and capable of providing early warning to the application regarding an impending reduction in the sending rate but also can be designed to have a smooth congestion response