具動態切換論域機制之Type-2模糊流量控制器設計

Type-2 Fuzzy Flow Controller Design with Dynamic Discourse Switching

陳昱甫(Yu-Fu Chen)

指導教授 ： 曾傳蘆

國立臺北科技大學/電資學院/電機工程系研究所/碩士(2009年)

摘要

本論文使用最近提出的控制系統傳輸協定(control system transfer protocol, CSTP)，此傳輸層協定具有改變傳輸速率功能，因此在應用層上設計適用於網路流量控制(flow control) 的Type-2模糊流量控制器(Type-2 fuzzy flow controller)。此控制器最大特色是引入Type-2模糊控制改善網路不確定因素(uncertainty)造成的影響，原理為使用基因演算法將控制器的比例因子最佳化，並利用網路端點對端點延遲時間(end to end delay, EED)及其變動量來推論傳送端傳輸速率，加上控制器設有『動態切換論域機制』，使傳送端能實現具有即時自我調節網路流量控制功能。
最後由網路模擬軟體NS-2(network simulation ver.2.27)比較傳輸控制協定(transmission control protocol, TCP)、使用者資料段協定(user datagram protocol, UDP)、即時傳輸協定RTP(real-time protocol)、應用層為Type-1模糊控制器(type-1 fuzzy flow controller)之CSTP協定與本論文所提出應用層為GA-Type-2模糊控制器之CSTP協定，結果顯示本論文提出之應用層為GA-Type-2模糊控制器之CSTP協定，綜合的性能指標效果較好。

關鍵字

流量控制 ； Type-2模糊控制 ； 基因演算法 ； 網路控制系統

並列摘要

The objective of this paper is to investigate the flow controller design for Internet. The adopted protocol is the currently proposed control system transfer protocol (CSTP) which has the ability of changing transmission rate. Consequently, we can design a type-2 fuzzy flow controller in the application layer together with CSTP to improve the transmission performance.
The benefit of this proposed controller is using type-2 fuzzy control to improve the network impact of uncertain factors. In addition, the scaling factors of the controller is optimized by genetic algorithm and the sending rate is adjusted according to the fuzzy inference of the end-to-end delay and the change of delay. Moreover, the controller provides the mechanism of dynamic regulation technique of discourse, to achieve real-time network-traffic dependent transmission rate regulation at sending node.
Finally, a network simulation software NS-2 is used to compare the performance of several protocols, including TCP, UDP, RTP, CSTP with conventional fuzzy flow controller, and CSTP with the proposed type-2 fuzzy flow controller. The simulation results show that the CSTP with the proposed GA-Type-2 fuzzy controller outperforms other protocols.

並列關鍵字

Flow control ； Type-2 fuzzy control ； Genetic algorithm ； Networked control system

參考文獻

[3]R. Oboe and P. Fiorini, “A design and control environment for Internet-based telerobotics,” International Journal of Robotics Research, vol. 17, Apr. 1998, pp. 433-449.

[4]E. Jammeh, M. Fleury, C. Hagras and M. Ghanbari, “All-IP network video streaming through interval type-2 fuzzy logic congestion control,” Proceedings of the IET International Conference on Intelligent Environments, Aug. 2008, pp. 452-457.

[6]Y. Wakita, S. Hirai, and K. Machida, “Intelligent monitoring system for limited communication path: telerobotic task execution over Internet,” Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems, vol. 2, Aug. 1995, pp. 104-109.

[7]L. A. Zadeh, “Fuzzy sets,” Information and Control, vol. 8, 1965, pp. 338-353.

[10]W. Xu, A. G. Qureshi, and K. W. Sarkies, “Novel TCP congestion control scheme and its performance evaluation,” IEE Communications, vol. 149, issue. 4, Aug. 2002, pp. 217-222.