City University of Hong Kong
DSpace
 

CityU Institutional Repository >
3_CityU Electronic Theses and Dissertations >
ETD - Dept. of Computer Science  >
CS - Doctor of Philosophy  >

Please use this identifier to cite or link to this item: http://hdl.handle.net/2031/4377

Title: End host multicast algorithms for overlay networks
Other Titles: Ji yu shang ceng wang luo de duan zhu ji zu bo suan fa de yan jiu
基於上層網絡的端主機組播算法的研究
Authors: Tu, Wanqing (塗萬青)
Department: Dept. of Computer Science
Degree: Doctor of Philosophy
Issue Date: 2005
Publisher: City University of Hong Kong
Subjects: Multicasting (Computer networks)
Notes: CityU Call Number: TK5105.887.T8 2005
Includes bibliographical references (leaves 181)
Thesis (Ph.D.)--City University of Hong Kong, 2005
xi, 181 leaves : ill. ; 30 cm.
Type: Thesis
Abstract: Multicast is a communication pattern in which a source host sends a message to a group of destination hosts. Compared with unicast, where multiple point-to-point connections must be set up to support communications among a group of users, multicast is more efficient for information distribution within a group of users. It is because each packet is replicated at the branching points leading to distinguished destinations, thus reducing the transmission load on the data sources and traffic load on the network links. Multicast is initially implemented in the network layer (i.e., IP layer) of OSI reference model that is called the IP multicast. The network carrying the IP multicast needs to incorporate new routing and forwarding mechanisms in addition to the existing unicast methods. However, the necessary functions needed to realize multicast are not adequately supported in the current networks. It incurs the serious deployment limitations of IP multicast. End host multicast holds the promise for realization of large scale Internet multicast services. An end host multicast is such a data structure that provides multicast functionality in the overlay networks through unicast connections among group members in the underlying layers. In contrasted to the IP multicast, the end host multicast has the following attractive features: (1) the concept of end host multicast enables the multicast deployment over heterogeneous networks without the need of universal network support; (2) end host multicast allows end hosts to directly create an overlay network session among themselves. This gives applications the flexibility to communicate without relying on service providers; (3) the flow control, congestion control and reliable delivery services in unicast can be easily employed in the end host multicast. The resource and the multicast implementation in the context of overlay networks are different from that in the convention networks that presents new challenges and new problems to solve. Our objectives in this dissertation are the efficient in terms of short multicast delay, scalable in terms of group size, fully distributed, multisource supporting and quality of service guaranteed end host multicast services. We propose a set of novel algorithms to construct the end host multicast and guarantee the quality of multicast services on the overlay networks. We present below the main contributions in this dissertation. Generally, two overlay topologies: tree and m-D mesh are for the end host multicast. We first analyze the fundamental issues relating to the efficient, scalable, distributed and multi-source supporting end host multicast services. In accordance with the analysis, we design a set of novel location-aware and capacity-aware algorithms to construct the end host multicast protocols to achieve the anticipative multicast services on top of the two overlay topologies respectively: the short delay end host multicast protocol (SDEM) in the context of tree topology in Chapter 3 and the distributed and scalable end host multicast protocol (DSM) on top of m-D mesh in Chapter 4. SDEM achieves the bottleneck-avoidance short multicast delay services through fully utilizing the underlying network properties and considering the end host heterogeneities in the output capacities; DSM employs the advantages of m-D mesh topology to design the multicast architecture and routing scheme that eliminates the cost of constructing trees but achieves the similar routing efficiency as the tree routing. The scalability of DSM is dominant when it is used in the multi-source group applications. For the quality of multicast services, we study the problem of decreasing the worstcase delay bound of end host multicast in Chapter 5. We propose a novel and simple adaptive control algorithm that employs a new (σ,ρ,λ) regulator at each group end host for effective delay control and resource utilization under the heavy network situations in the multi-group networks with high rate real-time flows. To our knowledge, it is the first work to incorporate the traffic regulator into the end host multicast to control the traffic output. A series of theoretical analysis and results on the worst-case delay bounds, the input rate threshold and the worst-case delay improvements is presented. In Chapter 6, we address the issue of balancing the decrease of end-to- end delay and the elimination of delay jitter caused by the fluctuation of network throughput during the packet transmission. A novel adaptive playback buffer (APB) is proposed to achieve the improved balance, namely, the continuous and real-time playback at the receivers, by utilizing the instantaneous network situations and playback performances.
Online Catalog Link: http://lib.cityu.edu.hk/record=b2107103
Appears in Collections:CS - Doctor of Philosophy

Files in This Item:

File Description SizeFormat
fulltext.html157 BHTMLView/Open
abstract.html157 BHTMLView/Open

Items in CityU IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

Valid XHTML 1.0!
DSpace Software © 2013 CityU Library - Send feedback to Library Systems
Privacy Policy · Copyright · Disclaimer