目前的DiffServ網路中，邊界(Edge)與核心(Core)路由器負責為不同的資料流分類成不同的Per-Hop Behaviors(PHBs)，並依照不同的PHBs給予不同的服務品質(Quality of Service)。而為了達成不同的QoS保證，則使用了各種不同的排程器。但IETF並無提出一個適當而有效的排程器來解決即時資料流(Real-time Traffic)所最關心的Jitter問題。
IETF在RFC中提出的EF資料流擁有低延遲(Low Latency)、低封包遺失率(Low Packet Loss Rate)與低封包延遲變化率(Jitter)等等的特性。即時資料流常被歸類為EF Flow，但處理即時資料流並非把封包快速送出即可，而是在一定的延遲時間內送出，過快或者過慢送出都不好。考慮到這樣的特性，我們將每一個Hop的Queuing Delay用一個值來代表，每個封包都有其預設的Per-Hop Queuing Delay。封包若在預設的Per-Hop Queuing Delay內送出完成，表示封包提早送出(具有Credit)，若是在預設時間外送出，表示此封包太晚送出(具有Deficit)。而Credit/Deficit可以隨著封包傳輸而夾帶在我們設計的 Credit/Deficit IPv6 Optional Header內，只要將封包內累積的Credit/Deficit消除，即能達到一個Constant Delay的傳輸功能，就能減小Jitter。為了設計一個具有Credit/Deficit的封包排程機制，我們選擇了Weighted Fair Queue(WFQ)為架構，並且加入了計算預估佇列延遲時間(Estimated Queuing Delay)與動態轉換等級(Dynamically Change Class)的功能，讓EF的封包能夠自由的在合適的佇列間跳動，以達到較小的封包延遲變化率，具有較為固定的Queuing Delay。
我們在Linux平台上先實作了WFQ、然後再實作本論文所提出的Credit/Deficit WFQ(CDWFQ)，並且以實驗結果證明我們的CDWFQ排程器可以讓EF的資料流擁有固定的佇列延遲、低封包遺失率與低延遲變化率。

In a DiffServ networks, edge and core router classify traffic flows into different PHBs and provide different QoS for the classified flows. In order to achieve satisfactory QoS guarantee, many packet schedulers were proposed. However IETF have not formally standardized an appropriate and effective packet scheduler to minimize the jitter for real-time traffic.
In RFC, EF flows are characterized with low-latency, low packet loss rate, and low jitter. Therefore, real-time traffic is often classified into EF flow. By considering the characteristics of real-time traffic, it is not appropriate to forward packets either too fast or too slow. Hence, in this Thesis, we propose a mechanism in which each packet is attached with its own per-hop queuing delay. If a packet is forwarded within its own per-hop queuing delay, we say the packet may arrive too early (credit accumulation). If a packet is forwarded beyond its own per-hop queuing delay, we say the packet has late arrival (deficit accumulation). The Credit/Deficit information can be stored in the IPv6 optional header so that it can pass through the whole networks. If we can minimize the Credit/Deficit, the jitter can be minimized too. Our design is based on a modified WFQ by adding functions such as estimated queuing delay and dynamic class changes. The dynamic class changes allow EF packets to switch among queues to achieve lower jitter and constant delay.
We first implement the traditional WFQ scheduler on Linux platform and then followed by the implementation of the Credit/Deficit WFQ (CDWFQ). The experimental results have shown that CDWFQ can provide nearly constant queuing delay, lower packet loss rate, and lower jitter for EF traffic flows.

第一章 導論……………………………………………………………………………………1
1.1 研究動機………………………………………………………………………………………1
1.2 研究及實作方法………………………………………………………………………………1
1.3 論文架構………………………………………………………………………………………3

第二章 差別式服務網路上的封包排程機制…………………………………………………4
2.1 差別式服務………………………………………………………………………………4
2.1.1 PHB(Per Hop Behavior)………………………………………………………5
2.1.2 差別式服務的欄位………………………………………………………………5
2.2 封包排程器(Packet Scheduler)………………………………………………6
2.2.1 封包排程機制的演進………………………………………………6
2.2.2 Weighted Round Robin(WRR)……………………………………7
2.2.3 Weighted Fair Queue(WFQ)………………………………………8
2.2.4 Deficit Round Robin(DRR)………………………………………9
2.3 減少Jitter的相關研究…………………………………………………………10
2.3.1 決對性的延遲保證………………………………………………11
2.3.2 以Round Robin為基礎的排程器的延遲討論……………………11
2.3.3 動態改變Quantum與Weight的方式………………………………12
2.3.4 動態切細循環的方式……………………………………………13
2.3.5 改進Weighted Fair Queue的方法………………………………14
2.3.5.1 Worst-Case Fair Weighted Fair Queue(WF2Q)………………………………………………………………………………………………………14
2.3.5.2 Leap Forward Virtual Clock(LFVC)………………………………………………………………………………………………………15
2.3.5.3 Least Time to Overflow的方式…………………15
2.4 新的Credit/Deficit概念………………………………………………………17
第三章 具有Credit/Deficit的動態調整排程器……………………………………………18
3.1 Credit/Deficit的架構……………………………………………………………………18
3.2 紀錄Credit/Deficit於封包內……………………………………………………………20
3.3 具有Credit/Deficit動態調整機制的排程器……………………………………………22
3.4 Credit/Deficit WFQ(CDWFQ)的架構圖……………………………………………………23
3.4.1 封包資料庫……………………………………………………………………24
3.4.2 封包抵達率的量測(Arrival Rate Measurement)…………………………25
3.4.3 預估延遲的計算(Estimated Delay Calculation)………………………25
3.4.4 動態改變等級(Dynamically Change Class)………………………………28
3.4.5 更新CD Optional Header……………………………………………………31
3.5 CDWFQ排程器的演算法………………………………………………………………………31

第四章 Linux上的實作與實驗結果分析………………………………………………………34
4.1 Linux路由器上的封包傳輸機制與排隊規則(Queuing Discipline)………34
4.2 Linux的IPv6封包接收順序與CD Optional Header的取出…………………35
4.3 Linux的IPv6封包發送順序與CD Optional Header的值入…………………37
4.3.1 UDP/ICMP封包的發送……………………………………………38
4.3.2 TCP封包的傳送過程……………………………………………39
4.3.3 CD Optional Header的植入……………………………………39
4.4 實驗拓樸………………………………………………………………………40
4.5 實驗一：測試預估的Delay……………………………………………………41
4.5.1 TCP Flow的預估Delay量測……………………………………42
4.5.2 UDP Flow的預估Delay量測……………………………………42
4.6 實驗二：TCP Flow在不同排程機制下的的Jitter比較……………………44
4.7 實驗三：UDP Flow在不同排程機制下的比較………………………………46
4-8 實驗四：不同的End-to-End Delay設定對於TCP Throughput影響………48
4-9 實驗結論………………………………………………………………………50

第五章 結論與未來發展………………………………………………………………………51
5.1 結論……………………………………………………………………………51
5.2 未來發展………………………………………………………………………52

Reference………………………………………………………………………………………53
索引……………………………………………………………………………………………56

[1] R. Braden, D. Clark, and S. Shenker, “Integrated Services in the Internet Architecture: an Overview”, RFC1633, June 1994.
[2] S. Blake, D. Black, M. Carlson, E. Davies, Z. Wang, and W. Weiss, “An Architecture for Differentiated Services”, RFC 2475, December 1998.
[3] K. Nichols, S. Blake, F. Baker, and D. Black, “Definition of the Differentiated Services Field (DS Field) in the IPv4 and IPv6 Headers”, RFC2474, December 1998.
[4] V. Jacobson, K. Nichols, and K. Poduri, “An Expedited Forwarding PHB”, RFC2598, June 1999.
[5] J. Heinanen, F. Baker, W. Weiss, and J. Wroclawski, “Assured Forwarding PHB Group”, RFC2597, June 1999.
[6] S. Deering and R. Hinden, “Internet Protocol, Version 6 (IPv6) Specification”, RFC2460, December 1998.
[7] J Nagle, “On Packet Switches with Infinite Storage”, IEEE Trans. On Communications, pp. 435-438, April 1987.
[8] H. Shimonishi, M. Yoshida, F. Ruixue, and H. Suzuki, “An Improvement of Weighted Round Robin Cell Scheduling in ATM Networks”, GLOBECOM '97, vol. 2, pp. 1119 –1123, 1997.
[9] Yoshihiro Ito,Shuji Tasaka, and Yutaka Ishibashi, “Variably Weight Round Robin Queueing for Core IP Routers”, IEEE International Conference on Performance, Computing, and Communications, pp. 159 –166, 2002.
[10] Taeck-Geun Kwon, Sook-Hyang Lee, and June-Kyung Rho, “ Scheduling Algorithm for Real-time Burst Traffic Using Dynamic Weighted Round Robin”, IEEE International Symposium on Circuits and Systems, vol. 6, pp. 506 –509, 1998.
[11] A. Demers, S. Keshav, and S. Shenker, “Analysis and Simulation of A Fair Queueing Algorithm”, SIGCOMM’89, vol. 19, no. 4, pp. 1-12, September 1989.
[12] M. Shreedhar and G. Varghese, “Efficient Fair Queuing Using Deficit Round-robin”, IEEE/ACM Transactions on Networking, vol. 4, Issue: 3, pp. 375 –385, June 1996.
[13] M. H. MacGregor and W. Shi, “Deficits For Bursty Latency-critical Flows: DRR++”, IEEE International Conference on Networks, pp. 287 –293, 2000.
[14] Yao Liang, “A Simple and Effective Scheduling Mechanism Using Minimized Cycle Round Robin”, IEEE International Conference on Communications, vol. 4, pp. 2384 –2388, 2002
[15] S.S. Kanhere and H.Sethu, “Fair, Efficient and Low-latency Packet Scheduling Using Nested Deficit Round Robin“, IEEE Workshop on High Performance Switching and Routing, pp. 6 –10, 2001.
[16] C.R.Jon Bennett and Hui Zhang, “WF2Q: Worst-case Fair Weighted Fair Queueing”, IEEE INFOCOM’96, pp.120-128, Mar. 1996.
[17] S. Suri, G.. Varghese, and G. Chandranmenon, “Leap Forward Virtual Clock: a New Fair Queueing Scheme with Guaranteed Delays and Throughput Fairness”, IEEE INFOCOM '97, vol. 2, pp. 557 -565, Apr. 1997.
[18] Mong-Fong Homg, Wei-Tsong Lee, Kuan-Rong Lee, and Yau-Hwang Kuo, “An Adaptive Approach to Weighed Fair Queue with QoS Enhanced on IP Network”, TENCON 2001, vol. 1, pp. 181 -186, 2001.
[19] N. G. Duffield. T. V. Lakshman, and D. Stiliadis, “On Adaptive Bandwidth Sharing with Rate Guarantees”, IEEE INFOCOM, pp. 122-130,1998.
[20] J.R.Gallardo and D. Makrakis, “Dynamic Predictive Weighted Fair Queueing for Differentiated Services”, IEEE International Conference on Communications, vol. 8, pp. 2380 –2384, 2001.
[21] J.R. Gallardo, D. Makrakis, and L. Orozco-Barbosa, “ Prediction of Alpha-stable Long-range Dependent Stochastic Processes”, Advances in Performance Analysis,vol 3(1), pp. 31-42, 2000.
[22] G. Malkin and R. Minnear,” RIPng for IPv6”, RFC 2080, January 1997.
[23] W. Almesberger, J. H. Salim, and A. Kuznetsov, “Differentiated Services on Linux”, Internet-Draft, June 1999.
[24] W. Almesberger, “Linux Network Traffic Control ― Implementation Overview”, EPFL ICA, February 4, 2001.
[25] Chuanxiong Guo and Shaoren Zheng , “Analysis and Evaluation of the TCP/IP Protocol Stack of Linux”, International Conference on Communication Technology Proceedings, vol 1, pp. 444 –453, Aug. 2000.
[26] Martin Devera aka devik, “HTB: Hierarchy Token Bucket”, http://luxik.cdi.cz/~devik/qos/htb/, May 2003

[27] A. Tirumala, F. Qin, J. Dugan, and J. Ferguson, ”Iperf Version 1.6.4,” http://dast.nlanr.net/Projects/Iperf/, May 2002.
[28] Christian Worm Mortensen, “Weight Round Robin, Linux Implementation”, http://wipl-wrr.sourceforge.net, Oct 2002.