透過您的圖書館登入
IP:3.15.221.136
  • 學位論文

TCP網路壅塞之模糊控制

Fuzzy Controller Design for TCP Congestion Control

指導教授 : 練光祐
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


隨著網際網路技術的興起,網路也因使用者大量增加而呈現壅塞的狀態。而在一些即時性的應用方面,其對於時間延遲現象將會十分的敏感。而時間延遲的主因,絕大多數取決於封包往返時間 (RTT) 的不確定性,而其中又以路由器上封包的排隊時間 (queue time) 影響最大。因此,我們主要著重於如何有效的控制路由器上的封包排隊長度,以有效的得知RTT的大小,也可降低封包遺失數及提升頻寬利用率,如此將對於網路品質的提升十分有幫助。 在本論文中,我們利用適用於非線性系統控制的 T-S 模糊模式 (Takagi-Sugeno fuzzy model),對上述 TCP 傳輸作 T-S 模糊模式化,針對此非線性系統利用模糊厜追蹤控制的理論設計模糊控制器以達到控制路由器上的封包排隊長度的目的。另外,利用座標轉換以穩態工作點作為新座標之原點,兼以模糊控制器之強健性,可以使控制器變得十分精簡,有效降低系統運算負荷。而在無線區域網路及多媒體傳輸方面,我們也嘗試配合其他協定以解決TCP傳輸效能不佳的問題。在章節中,我們詳述對閉迴路系統的穩定性分析以及呈現模擬結果及實驗數據。 最後我們使用網路模擬軟體 NS-2 來實現模糊控制器,而實驗結果也顯示,不論是RTT、封包排隊長度、頻寬利用率及丟包狀況,皆驗證了提出的模糊控制器皆具備強健性以及高效率的控制效果;此外,也提出了較為精簡的控制器形式,可實作於一般配備較低的路由器中。

並列摘要


Congestion control is an important issue in Internet communications where the con- gestion often occurs at routers. As a major method of congestion control, active queue management (AQM) regulates the queue length on the routers by adjusting the congestion window sizes of senders. In this thesis, we propose a Takagi -Sugeno (T-S) fuzzy con- troller for AQM with explicit congestion notification (ECN) to improve the performance of Internet communications. We use the queue length and incoming flow rate to measure the degree of congestion effect on Internet to regulate the queue length at a specified level. Linear matrix inequality (LMI) techniques are employed to obtain the feedback gains. Two schemes namely, fuzzy output regulation (FOR) and fuzzy stabilization (FS) are proposed. It is worthwhile that the FS AQM has a quite simple form and can be easily implemented. To proceed farther into verifying the proposed schemes, we make use of the network simulation (NS-2) in our experiment. Compared to other AQM schemes, the experiment results show that the proposed AQMs have better performance in terms of packet loss, throughput and the fluctuation of queue length.

參考文獻


[3] S. Floyd, and V. Jacobson, “Random Early Detection Gateways for Congestion Avoid- ance”, IEEE/ACM Transactions on Networking, Vol. 1, No. 4, August 1993.
[4] S. Athuraliya, S.H. Low, V. H. Li, and Q. Yin, “REM: Active Queue Management”, IEEE Network Magazine, Vol. 15, No. 3, May/June 2001.
[5] C. Hollot, V. Misra, D.Towsley, and W. Gong, “On Designing Improved Controllers for AQM Routers Supporting TCP Flows”, Proceedings IEEE INFOCOM’01, Alaska, USA, April 2001.
[7] V. Misra, W. Gong, and D. Towsley, “A Fluid-based Analysis of a Network of AQM Router Supporting TCP Flows with an Applicaton to RED”, Proceedings of ACM SIGCOMM’00, September 2000.
[8] P. Yan, Y. Gao, and H. Ozbay, “Variable Structure Control Approach to Active Queue Management for TCP with ECN”, accepted by IEEE Transactions on Control Systems Technology. ISCC 2003, Kemer-Antalya, Turkey, June 2003.

延伸閱讀


國際替代計量