HKUST Institutional Repository >
Electronic and Computer Engineering >
ECE Master Theses >
Please use this identifier to cite or link to this item:
|Title: ||Bandwidth allocation algorithms for file distribution networks and location-aware topology construction in peer-to-peer networks|
|Authors: ||Ou, Qi|
|Issue Date: ||2007 |
|Abstract: ||Peer-assisted applications such as the BitTorrent and PPStream have become more and more popular in today's Internet. They can provide the service of file distribution, multimedia streaming and video on demand. Many research problems have been raised due to the wide use of the peer-assisted applications. In this thesis, we study two problems in the peer-assisted networks: the bandwidth allocation problem for file distribution networks and the location-aware topology construction problem.
For the bandwidth allocation problem, we first formulate the minimum overall finishing time problem and the minimum average downloading time problem. We then provide a centralized bandwidth allocation algorithm that can achieve the minimum overall finishing time for simultaneous file distribution networks. For the dynamic networks, we present a simple algorithm to construct a network with link-level homogeneous property. In a link-level homogeneous network, all the links have the same uploading bandwidth so that the peers' downloading rates can be carefully controlled to be globally identical. With the proposed bandwidth allocation algorithms, optimal file distribution networks can be constructed. Extensive simulation results demonstrate that our protocol can greatly improve the performance of the file distribution networks.
Most of the peer-to-peer networks are constructed without considering the physical layer topology. This problem is often referred to as the topology mismatching problem. The topology mismatching problem results in low network efficiency due to the large amount of redundant traffic. To solve this problem, we propose an algorithm to construct a location-aware topology for the peer-to-peer networks. The simulation results show that our algorithm reduces the average neighbor distance significantly while it maintains the connectivity of the networks.|
|Description: ||Thesis (M.Phil.)--Hong Kong University of Science and Technology, 2007|
xiii, 67 leaves : ill. ; 30 cm
HKUST Call Number: Thesis ECED 2007 Ou
|Appears in Collections:||ECE Master Theses|
Files in This Item:
All items in this Repository are protected by copyright, with all rights reserved.