TCP是目前網際網路上傳送資料的主要協定，而其本身具有一些預防網路壅塞機制,如緩慢啟動，快速重送，壅塞預防，快速回覆等等，但這些機制往往解決了網路壅塞的問題，但也同時造成了網路的使用率下降。
傳統快速回覆的做法是先將擁塞視窗調整為原先的一半，以降低網路上流動的segment數量，進而解除擁塞的狀況。我們稱此種壅塞視窗調整法為傳統壅塞視窗調整法(TCWA)。可是傳統做法上並未仔細考慮為何會發生segment 遺失? 因為當傳送端偵測到segment遺失時，可能是單純的傳輸錯誤造成，而不是因為壅塞而導致的segment碰撞，如果不加以考慮這些狀況，只是一昧的將壅塞視窗降為一半，則會導致網路使用率下降。
在以往由於傳送端無法得知網路目前的狀態，所以其只能以最保守的方式,就是將segment遺失的發生都視為嚴重的擁塞，因而將壅塞視窗降為原先的一半，以求解決擁塞。
由於SACK 的出現，傳送端開始可以得知接收端目前的狀態以及有多少segment被網路丟棄，如果其判斷segment 遺失是由於傳輸錯誤，就不需要將壅塞視窗降為原先的一半，而可以隨著網路壅塞的狀況適當的調整壅塞視窗的大小，如此網路會一直維持在最大的使用率，我們稱此種方法為以選擇性回應為基礎的動態擁塞視窗調整法(SDCWA)。
此方法是主要是利用含有SACK不斷修正目前擁塞視窗的大小，使它可以盡可能的利用到網路所有的頻寬，以增加網路的效能。

TCP is a major protocol in the current Internet. It uses several mechanisms to prevent network congestion. These mechanisms can solve network congestion, however, they may result in throughput degradation.
In traditional Fast Recovery scheme, the sender reduces Congestion Window to a half. Therefore it results in the decrease of number of segments in the end-to-end pipe and solves the network congestion. We call it Traditional Congestion Window Adjustment, or TCWA for short.
Traditional processing did not examine what has caused the segment dropped. This may be due to the random loss or due to the sender being in congestion avoidance phase. In the past, Since a sender has no idea about network status using traditional control schemes, when it detects segment loss, it attributes the segment loss to serious congestion, and reduces the congestion window to a half.
Due to the development of TCP Sack, the sender is able to know the current network status such as the number of lost segments. According to these information, the sender can reduce its window to a proper size, so that high network utilization can be achieved.

1 Chapter 1 Introduction
1.1 Overview of TCP Sack and Sacked Dynamic Congestion Window Adjustment1.2 Organization
2 Chapter 2 Background
2.1 Transmit Control Protocol
2.2 TCP Features
2.2.1 Slow Start
2.2.2 Fast Retransmit
2.2.3 Fast Recovery
2.2.4 Congestion Avoidance
2.3 TCP Versions
2.3.1 TCP Tahoe
2.3.2 TCP Reno
2.3.3 TCP New Reno
2.3.4 TCP Sack
3 Chapter 3 Sack Based Dynamic Congestion Window Adjustment
3.1 Why do we need Sack Based Dynamic Congestion Window Adjustment
3.2 Sack Based Dynamic Congestion Window Adjustment
3.2.1 Analysis of TCP Sack Fast Recovery phase behavior
3.2.2 Sack Based Dynamic Congestion Window Adjustment
3.2.3 Comparisons of TCWA and SDCWA
4 Chapter 4 The Simulation Models and Results
4.1 Configuration and Assumption
4.2 Simulation results and Analysis
4.2.1 Simulation1
4.2.2 Simulation2
4.2.3 Simulation3
4.2.4 Simulation4
5 Chapter 5 Conclusion and Feature Works
5.1 Conclusion
5.2 Future Works

[1] V.Jacobson, "Congestion Avoidance and Control," Proceedings of SIGCOMM'88 Symposium, pp.314-32, August 1988
[2] Kevin Fall, Sally Folyd, "Simulation-based Comparisons of Tahoe, Reno and SACK TCP"
[3] Sally Floyd, "Issues of TCP with SACK"
[4] M.Mathis, J. Madhavi, S. Floyd, A. Romanow, "TCP Selective Acknowledgement Option," Internet RFC 2018, October 1996
[5] W.Stevens, "TCP Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery Algorithms," internet RFC 2001, January 1997
[6] Janey C. Hoe, "Startup-up Dynamics of TCP's Congestion Control and Avoidance Schemes," MS Thesis, Massachusetts Institute of Technology, June 1995
[7] Mark Allman, Chris Hayes, Hans Kruse, Shawn Ostermann, "TCP Performance over Satellite Links," Proc. 5th International Conference on Telecommunications System, 1997
[8] Kedarnath Poduri and Kathleen Nichols. "Simulation Studies of Increased Initial TCP Window Size," June 1998.
[9] Lawrence S. Brakmo, Sean W. O'Malley, and Larry L. Peterson. "TCP Vegas: New Techniques for Congestion Detection and Avoidance," Proceedings of ACM SIGCOMM '94, August 1994
[10] S. Folyd. "SACK TCP: The Sender Congestion Control Algorithms for the implementation "sack1" in LBNL's "ns" simulator.," Technical report, Mar. 1996. Presentation to the TCP Large Windows Working Group of the IETF, March 7,1996
[11] W. Stevens. TCP/IP Illustrated, volume 1. Addison-Wesley, Reading MA 1994.
[12] Sally Floyd, "TCP and Successive Fast Retransmits", February 1995. Obtain via ftp://ftp.ee.lbl.gov/papers/fastretrans.ps
[13] Peter B.Danzig and Sugih Jamin. "A library of TCP/IP traffic characteristic," Technical Report TR-SYS-91-01, USC Networking and Distributed Systems Laboratory, October, 1991. Obtain via ftp://catarina.usc.edu/pub/jamin/tcplib
[14] V. Jacobson, R.Braden, and D. Borman. "TCP Extensions for High performance," May 1992. Reguest for Comments 1323.
[15] V. Jacobson. R. Braden. "TCP extensions for long-delay paths," October 1988, Request for Comments 1072