使用模糊邏輯演算法提升802.11e WLANs QoS 效率之研究

Improving the QoS Performance of EDCA in IEEE 802.11e WLANs Using Fuzzy Set Theory

王佳鈴(Chia Ling Wang)

指導教授 ： 黃依賢

元智大學/資訊學院/資訊工程學系/碩士(2004年)

摘要

以往無線網路主要定義封包資料的傳輸，一般資料的傳輸允許在網路擁塞的情況下遺失資料，透過重送的機制來把資料傳遞完成，可是這樣的方式並不能充分滿足即時(Real-Time)多媒體影音服務傳輸。
所以IEEE 802.11e定義了QoS的機制，提供影音資料較高的傳輸優先權，讓使用者更流暢地使用無線網路影音服務。但IEEE802.11e存取失敗時，競爭視窗值(contention window, CW)呈指數型態的成長，導致CW值成長快速的問題，並不符合資料實際使用狀態。因此本文期望改善的情形有兩種，其一：當傳輸失敗(發生碰撞)時，依據網路變因，動態調整CW值，使CW值合理的成長。其二：當某節點多次傳輸失敗重傳至成功時，因為傳輸失敗時CW值成長到最大CW值，而重傳至成功時，CW值會設定為原先 值，此狀況下可能會發生突然的碰撞(bursty collisions)，所以本文中傳輸成功時CW值不直接設成 值，而是使用本文演算法慢慢回到 值。藉由此兩種狀況的改善，期望提供高優先權服務品質，並儘量不影響低優先權的服務需求，降低封包延遲。本文利用模糊理論(Fuzzy Theory)，提出fuzzy CW allocation control (FCWAC)的演算法去控制backoff function之CW值。FCWAC使用工作站之佇列負载程度(queue length load)、封包在佇列內有效等待時間(available waiting time)及封包傳送的遺失機率(lose probability)之模糊集合(Fuzzy set)與歸屬函數(membership function)來調整CW值。在論文中，將建置模擬環境所需的參數、條件等，以實驗進行模擬、效益分析，經由delay time，drop rate，throughput模擬結果與相關圖表來評估演算法的執行效能。

關鍵字

模糊理論 ； 效能分析

並列摘要

IEEE 802.11 wireless LAN (WLAN) is the most widely used WLAN standard nowadays, but it cannot provide QoS support for the demanding multimedia applications. Instead, the IEEE 802.11e has defined the QoS support to a priority scheme by differentiation the inter-frame space and the initial window size for WLAN multimedia applications. Two important problems resided in many research: one is that the contention window (CW) doubled after each unsuccessful transmission attempt until a predefined maximum is reached, which increase the packet delay time due to the CW value grows too fast; the second is that CW will be back to CWmin abruptly after successful transmission regardless of the network conditions, which possibly have the sudden collision (so called bursty collision). This paper proposes a fuzzy dynamic CW Algorithm to dynamically adjust the CW value to alleviate the two problems. Three input parameters in the fuzzy algorithm which are data queue length, packet loss probability and packet waiting time. The system performance is analyzed and compared in terms of the packet delay time, packet drop probability and throughput.

並列關鍵字

IEEE802.11e ； QoS ； contention window ； Fuzzy Theory ； Performance

參考文獻

[12] D. Slonowsky, Q. Jiang and R. Srinivasan, “A Fuzzy Dynamic Bandwidth Re-Allocator”, IEEE Electrical and Computer Engineering, Vol. 2, pp. 815-822, May 2002.

[4] S. Choi, J. Prado, S. Shankar and S. Mangold, “IEEE 802.11e Contention-Based Channel Access (EDCF) Performance Evaluation”, IEEE International Conference, Vol. 2, pp. 1151-1156, May, 2003.

[5] Y. Xiao, “Enhanced DCF of IEEE 802.11e to Support QoS”, IEEE Wireless Communications and Networking, Vol. 2, pp. 1291-1296, March 2003.

[6] G.W. Wong, R.W. Donaldson, “Improving the QoS performance of EDCF in IEEE 802.11e wireless LANs”, IEEE Computers and signal Processing, Vol. 1, pp. 392-396, Aug. 2003.

[8] K. Xu, Q. Wang and H. Hassanein, “Performance Analysis of Differentiated QoS by IEEE 802.11e Enhanced Distributed Coordination Function (EDCF) in WLAN”, IEEE Global Telecommunications Conference, Vol. 2, pp. 1048-1053, Dec.2003.