臺灣博碩士論文加值系統

網路服務日漸發達，各家網路服務業者也開始提供各種不同的網路服務，包含語音電話、視訊會議... 等。為了獲得正常的服務，這些服務都必須要達到需求的服務品質(Quality of Service) ，因此網路使用者會與網路服務業者簽訂服務等級契約(Service Level Agreement)，以保證網路服務能達到網路使用者的需求。由於網路服務的使用不一定會再同一網路服務業者內，以單一網路管理人員的觀點，無法得知這些跨網域服務的服務狀況，而使使用者的權利被忽略。
本論文提出以點對點(Peer-to-Peer)為基礎之網路連線服務監測系統，以網路使用者的觀點出發，利用分散在網路上的大量使用者，作全面性的網路連線服務監測，使一般的網路使用者能夠利用系統判斷其網路是否達到與網路服務業者訂立之服務等級契約，保證使用者之權利。本系統使用點對點之架構，使監測系統擁有彈性及強健度，使其他同一區域之使用者能代替離線使用者測量，維持監測資料之連續性，且利用點對點架構使用者分散之特性，增加監測的範圍，且不須經由網路服務業者更動網路設備的設定。每個監測目標使用頻寬約為2696.536bps，對ADSL來說足以使用，可避免從單點監測時，若監測封包被阻擋就無法取得服務資訊。本系統分兩種型態，一為安裝於網路使用者之監測代理員(Monitor Agent)，監測代理員之間互相監測取得連線服務資訊；另一為監測資訊收集者(Monitor Information Collector)，收集監測代理員所取得之連線服務資訊，以供查詢。

Network services are more and more popular, and every Internet Service Provider (ISP) begins to offer all kinds of network services, including Voice over IP (VoIP), video conference and so on. In order to receive the normal service, these services must meet Quality of Service (QoS) requirements demanded by customers. To define normal services, the customers will sign Service Level Agreements (SLA) with ISPs to ensure ISP can reach QoS, which customers demand. Because the customers may use the network services in different ISPs, network managers are not able to know network service status across network domains, and the right of the customers is ignored.
This thesis proposes a P2P-based QoS Monitoring System (QMS). It considers the customers view, gathers a large number of users that distributed in the network, and makes comprehensive network connection service monitoring, so that common customers can use this system to determine whether its network meets SLA signed with the ISP, and guarantees customers'' right. The proposed system uses P2P architecture to improve flexibility and robustness. Even if a user leaves this system, another user at the same network domain can replace to maintain QoS information of monitoring continuously. In addition, the feature of the P2P architecture that users are distributed everywhere is used to increase the range of QoS monitoring and ISPs need not to change configurations of network devices. Every monitored destination uses 2696.536 bps, it is enough for ADSL. The proposed system can avoid problem that user cannot get QoS information because ISPs block packets from some domains in using firewall or network device configuration. The proposed system can get QoS information from other nodes. This system is divided into the two kinds of roles, One is the Monitor Agent, which is installed in customers’ computers, and monitors each other to get network connection information between two Monitor Agents; Another is the Monitor Information Collector, which collects network connection information form the Monitor Agent and provides information to all customers who install this system.

Chapter 1 Introduction 1
1.1 Background 1
1.2 Motivations 3
1.3 Goals 5
1.4 Organization 6
Chapter 2 Related Work 6
2.1 Service Level Agreement 6
2.2 Peer-to-Peer Network 7
2.3 Distributed Hash Table and Chord 8
2.4 WHOIS 9
2.5 PeerCQ 10
2.6 ManP2P 11
Chapter 3 System Design 12
3.1 System Architecture 12
3.2 Monitor Agent 15
3.2.1 Graphic User Interface of Monitor Agent 16
3.2.2 Distributed Hash Table Module of Monitor Agent 17
3.2.3 Monitor Module of Monitor Agent 18
3.2.4 Node Management Module of Monitor Agent 19
3.2.5 Message Handling of Monitor Agent 20
3.2.6 Log Module of Monitor Agent 21
3.3 Monitor Information Collector 22
3.3.1 Graphic User Interface of Monitor Information Collector 24
3.3.2 Distributed Hash Table Module of Monitor Information Collector 25
3.3.3 Monitor Module of Monitor Information Collector 26
3.3.4 Node Management Module of Monitor Information Collector 27
3.3.5 Request Handling of Monitor Information Collector 28
3.3.6 Log Module of Monitor Information Collector 29
3.4 Monitor Information Design 30
3.5 Decide Master MIC and Schedule Tasks Algorithm 31
3.6 Message Flows in MAs and MICs 34
3.6.1 Message Flows in MAs 34
3.6.2 Message Flows in MICs 37
3.6.3 Message Flows among MAs and MICs 39
Chapter 4 Implementation 41
4.1 System Equipments and Environment 41
4.2 Achievements of proposed systems 44
4.2.1 Monitor Agent 44
4.2.2 Monitor Information Collector 47
Chapter 5 Experiments 51
5.1 Experiment 1 Functionality verification of all designed modules 51
5.2 Experiment 2 The impact of the number of MICs on the initial time 55
5.3 Experiment 3 Functionality verification of detecting disconnection for MA node management 60
5.4 Experiment 4 Functionality verification of detecting disconnection for MIC node management 63
5.5 Experiment 5 The impact of the number of MA and MIC subsystems on bandwidth consumptions 67
5.6 Experiment 6 Comparison between Monitoring from single Node and QMS 69
5.7 Experiment 7 Functionality verification of the proposed system on the Internet 72
5.8 Experiment 8 The impact of the background traffic on delay time for robustness 74
Chapter 6 Discussions 77
6.1 General Cases 77
6.2 Special Cases 79
6.2.1 ICMP denied 79
6.2.2 Mobile IP, VPN, Tunneling 80
6.2.3 Fabrication 81
Chapter 7 Conclusions and Future Work 82
7.1 Conclusions 82
7.2 Future Work 83
Reference ………………………………………………………………………..84

[1]S. Shenker, C. Partridge, R. Guerin, “Specification of Guaranteed Quality of Service”, RFC 2212, September 1997.
[2]K White, “Definitions of Managed Objects for Service Level Agreement Performance Monitoring”, RFC 2758, February 2000.
[3]ITU-T, “Principles for a Telecommunications Management Network (TMN)”, Rec. M.3010, 1996.
[4]Stephanos Androutsellis-Theotokis and Diomidis Spinellis. “A survey of peer-to-peer content distribution technologies”, ACM Computing Surveys, 36(4):335–371, December 2004.
[5]Lakshminarayanan Subramanian, Ion Stoica, Hari Balakrishnan, and Randy Katz, “OverQoS: An Overlay Based Architecture for Enhancing Internet QoS”, Proc. 1st Symposium on Networked Systems Design and Implementation (NSDI), San Francisco, CA, March 2004.
[6]Peer-to-Peer, http//en.wikidepia.org/wiki/Peer-to-Peer
[7]Hari Balakrishnan, M. Frans Kaashoek, David Karger, Robert Morris, and Ion Stoica. “Looking up data in P2P systems”, Communications of the ACM, February 2003.
[8]Ion Stoica, Robert Morris, David Karger, M. Frans Kaashoek, and Hari Balakrishnan, “Chord: A Scalable Peer-to-peer Lookup Service for Internet Applications”, Proceedings of ACM SIGCOMM 2001, San Deigo, CA, August 2001
[9]The Chord Project, http://pdos.csail.mit.edu/chord/
[10]L. Daigle, “WHOIS Protocol Specification”, RFC 3912, September 2004.
[11]Gedik, B., Liu, L., “A scalable peer-to-peer architecture for distributed information monitoring applications”, Computers, IEEE Transactions on
Volume 54, Issue 6, pp.767 – 782, Jun 2005.
[12]Granville, L.Z.da Rosa, D.M. Panisson, A. Melchiors, C. Almeida, M.J.B., “Managing Computer Networks Using Peer-to-Peer Technologies”, Communications Magazine, IEEE, Volume 43, Issue 10, pp. 62- 68, October 2005
[13]Open chord, https://www.lspi.wiai.uni-bamberg.de/dmsg/software/open_chord/
[14]D-ITG, http://www.grid.unina.it/software/ITG/
[15]Pchar, http://www.kitchenlab.org/www/bmah/Software/pchar/