Collaborative interest management for peer-to-peer massively multiplayer online games

In the past few decades, large-scale networked virtual environments~(NVEs) such as massively multiplayer online games~(MMOGs) have emerged, and it may involve thousands or even millions of concurrent players all over the world. The client-server architecture has the problem of scalability due to the...

Full description

Saved in:
Bibliographic Details
Main Author: Liu, Cheng
Other Authors: Cai Wentong
Format: Theses and Dissertations
Language:English
Published: 2015
Subjects:
Online Access:http://hdl.handle.net/10356/62534
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-62534
record_format dspace
spelling sg-ntu-dr.10356-625342023-03-04T00:33:57Z Collaborative interest management for peer-to-peer massively multiplayer online games Liu, Cheng Cai Wentong School of Computer Engineering Parallel and Distributed Computing Centre DRNTU::Engineering::Electrical and electronic engineering In the past few decades, large-scale networked virtual environments~(NVEs) such as massively multiplayer online games~(MMOGs) have emerged, and it may involve thousands or even millions of concurrent players all over the world. The client-server architecture has the problem of scalability due to the resource limitation on the centralized servers. The Peer-to-Peer~(P2P) architecture can improve the scalability by avoiding the limitation of servers and is thus a more suitable architecture for large-scale NVEs. Interest Management~(IM) is a fundamental issue in NVEs. It aims at disseminating the state of the virtual environment to the participating players according to their actual needs. The IM mechanism in the P2P architecture for large-scale NVEs is the focus of this thesis. Directly applying the existing IM mechanisms in the P2P NVEs may result in some peers having high communication or computational workload, or incur high communication cost due to global information update among peers. A collaborative interest management mechanism~(i.e., the CIM mechanism) is first proposed in the thesis to address the problems of applying the existing IM mechanisms to the P2P NVEs. The proposed CIM mechanism can be implemented easily by extending the existing IM mechanisms. The CIM mechanism is evaluated and the experimental results show that it has performance advantages over the existing IM mechanisms in terms of the communication cost and the workload of peers. The proposed CIM mechanism works on the virtual environment~(VE) which is divided into cells. How the VE is divided has influence on the performance of the CIM mechanism and the issue of VE division is then investigated. A performance model is introduced to analyze the communication cost and failure rate of the CIM mechanism and two algorithms are proposed to generate the division of VE in order to minimize the communication cost and failure rate of the CIM mechanism. The network delay among peers is highly related to the consistency issue of P2P NVEs. Finally, how the network delay can affect the proposed CIM mechanism is investigated in this thesis. We define a player to be a {\em consistent subscriber} when it meets the consistency requirement in the virtual environment and a mathematical model is introduced to analyze how the number of {\em consistent subscriber} changes over time. A superpeer-assisted mechanism is proposed and evaluated to maintain the consistency requirement for the CIM mechanism. In summary, this thesis focuses on the IM mechanism for large-scale P2P NVEs, as well as the related issues of IM mechanism, including VE division and network delay. The experimental results show that the proposed approaches in this thesis successfully reduce the communication cost of IM mechanisms for the P2P NVEs. Since IM is a fundamental issue in NVEs, the proposed approaches in this thesis can serve as the basis for further research in P2P NVEs. Doctor of Philosophy (SCE) 2015-04-14T08:28:37Z 2015-04-14T08:28:37Z 2015 2015 Thesis http://hdl.handle.net/10356/62534 en 167 p. application/pdf
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic DRNTU::Engineering::Electrical and electronic engineering
spellingShingle DRNTU::Engineering::Electrical and electronic engineering
Liu, Cheng
Collaborative interest management for peer-to-peer massively multiplayer online games
description In the past few decades, large-scale networked virtual environments~(NVEs) such as massively multiplayer online games~(MMOGs) have emerged, and it may involve thousands or even millions of concurrent players all over the world. The client-server architecture has the problem of scalability due to the resource limitation on the centralized servers. The Peer-to-Peer~(P2P) architecture can improve the scalability by avoiding the limitation of servers and is thus a more suitable architecture for large-scale NVEs. Interest Management~(IM) is a fundamental issue in NVEs. It aims at disseminating the state of the virtual environment to the participating players according to their actual needs. The IM mechanism in the P2P architecture for large-scale NVEs is the focus of this thesis. Directly applying the existing IM mechanisms in the P2P NVEs may result in some peers having high communication or computational workload, or incur high communication cost due to global information update among peers. A collaborative interest management mechanism~(i.e., the CIM mechanism) is first proposed in the thesis to address the problems of applying the existing IM mechanisms to the P2P NVEs. The proposed CIM mechanism can be implemented easily by extending the existing IM mechanisms. The CIM mechanism is evaluated and the experimental results show that it has performance advantages over the existing IM mechanisms in terms of the communication cost and the workload of peers. The proposed CIM mechanism works on the virtual environment~(VE) which is divided into cells. How the VE is divided has influence on the performance of the CIM mechanism and the issue of VE division is then investigated. A performance model is introduced to analyze the communication cost and failure rate of the CIM mechanism and two algorithms are proposed to generate the division of VE in order to minimize the communication cost and failure rate of the CIM mechanism. The network delay among peers is highly related to the consistency issue of P2P NVEs. Finally, how the network delay can affect the proposed CIM mechanism is investigated in this thesis. We define a player to be a {\em consistent subscriber} when it meets the consistency requirement in the virtual environment and a mathematical model is introduced to analyze how the number of {\em consistent subscriber} changes over time. A superpeer-assisted mechanism is proposed and evaluated to maintain the consistency requirement for the CIM mechanism. In summary, this thesis focuses on the IM mechanism for large-scale P2P NVEs, as well as the related issues of IM mechanism, including VE division and network delay. The experimental results show that the proposed approaches in this thesis successfully reduce the communication cost of IM mechanisms for the P2P NVEs. Since IM is a fundamental issue in NVEs, the proposed approaches in this thesis can serve as the basis for further research in P2P NVEs.
author2 Cai Wentong
author_facet Cai Wentong
Liu, Cheng
format Theses and Dissertations
author Liu, Cheng
author_sort Liu, Cheng
title Collaborative interest management for peer-to-peer massively multiplayer online games
title_short Collaborative interest management for peer-to-peer massively multiplayer online games
title_full Collaborative interest management for peer-to-peer massively multiplayer online games
title_fullStr Collaborative interest management for peer-to-peer massively multiplayer online games
title_full_unstemmed Collaborative interest management for peer-to-peer massively multiplayer online games
title_sort collaborative interest management for peer-to-peer massively multiplayer online games
publishDate 2015
url http://hdl.handle.net/10356/62534
_version_ 1759857881631424512