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

Towards automatic fairness for IP network applications

Aldabbagh, G.A.; (2011) Towards automatic fairness for IP network applications. Doctoral thesis , UCL (University College London).

Full text not available from this repository.

Abstract

The Transmission Control Protocol (TCP) carries the majority of today’s Internet traffic due to the success of the responsive congestion control mechanisms. However, the number of streaming media applications has grown tremendously during the past several years. Such applications do not use the TCP protocol, this implies that they rarely provide end-to-end congestion control (i.e. they are unresponsive) in a TCP-friendly manner and that they do not share the available bandwidth fairly with applications built on TCP. Additionally, it has been observed that short-lived flows (SLF) are at a disadvantage when competing against long-lived flows (LLF). This dissertation presents the fair bandwidth allocation architectural framework to provide fair allocation of network resources for competing flows traversing the network. The major contribution of this work is the design and implementation of a router-based Active Queue Management (AQM) scheme, Fair Early Drop (FED) which provides fair distribution of network bandwidth amongst competing responsive and unresponsive flows and LLFs and SLFs. A traffic classification module is also developed to identify LLFs from SLFs to enable FED to preferentially treat SLFs over LLF. SLFs are allowed to pass through the network without dropping any packets from them.

Type: Thesis (Doctoral)
Title: Towards automatic fairness for IP network applications
Language: English
Additional information: Permission for digitisation not received
UCL classification: UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Electronic and Electrical Eng
URI: https://discovery.ucl.ac.uk/id/eprint/1301767
Downloads since deposit
0Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item