UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Routing and congestion control in datagram networks

Wang, Zheng; (1992) Routing and congestion control in datagram networks. Doctoral thesis (Ph.D), UCL (University College London). Green open access

[img]
Preview
Text
Routing_and_congestion_control.pdf

Download (4MB) | Preview

Abstract

Routing and congestion control can be viewed as two closely related processes in a resource control system which manages the network resources. One of the basic problems is to adapt to changing conditions. Our analysis shows that inadequate information and delayed feedback can cause oscillation and instability. In this dissertation we examine the problems in routing and congestion control in a datagram environment and propose a number of solutions. The essence of the ideas is to deal with momentary fluctuations and average steady trends separately. The behavior of shorest-path routing algorithms is investigated. A new approach for dynamic routing based on the concept of decentralized decision making is introduced. A dynamic routing algorithm based on this new approach. Shortest Path First with Emergency Exits (SPF-EE), is developed, which eliminates many of the problems that the Shortest Path First (SPF) algorithm exhibits under heavy and dynamic traffic. The SPF- EE algorithm allows local and temporary alteration of routes without global route updating. Simulation shows that the SPF-EE algorithm achieves substantial improvement over the SPF algorithm. A new traffic adjustment algorithm called synchronized traffic adjustment (STA) and three information feedback schemes are proposed. The STA algorithm attempts to establish the sharing of the resources only when there are significant traffic changes and to maintain fairness by synchronized adjustment. Three congestion control schemes, based the STA approach and the proposed binary and quantitative feedback schemes, are presented. Extensive simulation shows that the STA algorithm eliminates the oscillation problems in the additive increase and multiplicative decrease (AIMD) algorithm and converges more quickly.

Type: Thesis (Doctoral)
Qualification: Ph.D
Title: Routing and congestion control in datagram networks
Open access status: An open access version is available from UCL Discovery
Language: English
Additional information: Thesis digitised by ProQuest.
Keywords: Applied sciences; Network congestion
URI: https://discovery.ucl.ac.uk/id/eprint/10107561
Downloads since deposit
78Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item