Responsive image
博碩士論文 etd-0703102-221932 詳細資訊
Title page for etd-0703102-221932
論文名稱
Title
FTCP, Csnoop - 在有線及無線網路上兩個TCP新技術之研究
FTCP, Csnoop - Two Novel Strategies for TCP over Wired and Wireless Network
系所名稱
Department
畢業學年期
Year, semester
語文別
Language
學位類別
Degree
頁數
Number of pages
66
研究生
Author
指導教授
Advisor
召集委員
Convenor
口試委員
Advisory Committee
口試日期
Date of Exam
2002-06-21
繳交日期
Date of Submission
2002-07-03
關鍵字
Keywords
傳輸控制協定、公平性、網路、壅塞控制、公平傳輸控制協定、窺探法、連續窺探法、無線網路
TCP, Fairness, Congestion Control, Network, Csnoop, Snoop, Wireless Networks, FTCP
統計
Statistics
本論文已被瀏覽 5660 次,被下載 3117
The thesis/dissertation has been browsed 5660 times, has been downloaded 3117 times.
中文摘要
中 文 摘 要
一條TCP連線的傳送速度取決於壅塞視窗(congestion window)的大小。壅塞視窗會在每接收到一個ACK時增加。這導致了TCP對於長往返時間(round-trip-time)的連線有偏差的待遇。為了增進TCP的公平性,我們提出了一個新的方法FTCP (Fair TCP)。跟TCP不同的是,在FTCP壅塞避免狀態(congestion avoidance state)時,FTCP會將本身的RTT跟標準RTT(standard RTT)做比較以調整每收到一個ACK時,cwnd的增加量。因此FTCP可以保持不同RTT的連線間有相同的傳送速度增加速率。在FTCP傳送逾時狀態(timeout state)時,FTCP計算出當標準連線(standard connection)到達其ssthresh的當時,這條連線的cwnd與原本cwnd / 2的差值,用來調整ssthresh的值以消除不同RTT的連線間在經過slow start state之後傳送速度的差異。FTCP可以相當程度的改進不同RTT的連線間頻寬分配不公平的情況。
因為不必要的壅塞控制,無法有效處理大量集中出現錯誤,以及長時間的延遲造成cwnd的回復速度減慢,這幾項原因造成了在無線網路中的TCP連線效能極差。在已經被提出來的改進方法中,Snoop把基地台當成一個用來暫存尚未被確認封包群的前端代理點。當無線網路發生錯誤時,Snoop會從基地台重傳遺失的封包以代替原本的TCP由發送端重傳。而且Snoop會把因為無線網路的錯誤而產生的重覆ACK丟棄以避免發送端發動壅塞控制。Snoop採用了與原本TCP相同的重傳方式。一個RTT只重傳一個遺失的封包。比起TCP,Snoop可以更快速的修復遺失的封包,也可以容忍更高的位元錯誤率(bit error rate)。 但是Snoop沒有真正解決原本TCP中因為多個封包遺失而產生的低效能問題。 我們提出了一個由Snoop延伸出來的新方法,Csnoop (continuous snoop)。 當大量且集中的錯誤在無線網路發生時,Csnoop在第一個RTT首先重傳一個遺失的封包,然後透過計算到達基地台的ACK數目來推算出遺失封包的個數。接著在下一個RTT中連續重傳遺失個數的封包。當無線網路的傳送逾時(local timeout)發生時,Csnoop推斷所有的傳送出去的封包都已經遺失,因此把在基地台中所有暫存的封包群馬上全部重傳。模擬結果顯示Csnoop可以達到比TCP與Snoop更好的傳送速度,尤其是當無線網路的頻道狀態極差時。再者,Csnoop在基地台中所需要暫存封包的空間也比Snoop少。
Abstract
Abstract
The throughput of a TCP connection is decided by the size of the congestion window. And cwnd increases when an acknowledgement arrives. It leads to that TCP has a bias against connections with long round-trip-time. For enhancing the fairness of TCP, we proposed a new scheme FTCP (Fair TCP). Unlike TCP, in FTCP congestion avoidance state, it compares its RTT with the standard RTT to adjust the increase amount of cwnd when an ACK arrives TCP sender. Therefore FTCP can keep the throughput increase rate of connections with different RTTs be the same. When FTCP enters timeout state, it sets appropriate slow start threshold by calculating the difference value of cwnd / 2 and the cwnd while standard connection achieves ssthresh. So that FTCP can eliminate the difference of throughput between connections with different RTT while leaving the slow start state. FTCP significantly improves the unfair bandwidth distribution between connections with different RTT.
TCP connections over wireless links perform badly because of the unnecessary congestion control, inefficiency to burst packet loss, and long delay to slow down the cwnd recovery time. In proposed schemes, Snoop takes BS as a pivot point to cache the unacknowledged TCP packets. When errors occur in wireless link, Snoop retransmits the packets locally from BS instead of retransmitting these packets from sender. And Snoop shields off the duplicate ACKs caused by wireless errors to avoid sender triggering unnecessary congestion control. But Snoop adopts same retransmission style as TCP. It only retransmits one packet per continuous duplicate ACKs. Snoop recovers error packets more quickly and tolerates higher BER than TCP. But Snoop doesn’t really solve the degraded performance problem of multiple errors of TCP. When the channel is the in a very bad quality, Snoop still performs badly. We proposed a new scheme, Csnoop (continuous snoop), extended from Snoop. When bursty errors happen in the wireless links, Csnoop retransmits one lost packets from the BS in first RTT and counts the number of ACKs that arrives BS to calculate the number of lost packets. And Csnoop retransmits these lost packets continuously. When local timeout happens, Csnoop infers that all packets were dropped and retransmits all packets cached in the buffer. Simulations show that Csnoop achieves better throughput compared to Snoop and TCP, especially for bad quality wireless links. Furthermore, Csnoop needs less buffer size to cache the unacknowledged packets at the base station than Snoop.
目次 Table of Contents
Content Page
中文摘要………………………………………………………………………. i
Abstract…………………………………………………………………………. iii
Content…………………………………………………………………………. v
List of Figures………………………………………………………………….. viii
List of Tables…………………………………………………………………... x
1. Introduction…………………………………………………………………. 1
2. Related Works…………………………………………………………….…. 4
2.1 TCP Overview…………………………………………………………. 4
2.2 TCP Congestion Control Algorithm…………………………………… 5
TCP Slow Start State…………………………………………………. 5
TCP Congestion Avoidance State…………………………………….. 6
TCP Fast Retransmission and Fast Recovery State…………………... 6
TCP Timeout State……………………………………………………. 8
An Example of TCP Cwnd…………………………………………… 8
2.3 The Problems of TCP………………………………………………….. 10
Inefficient to Multiple Packet Loss…………………………………... 10
Unfairness of Long Propagation Delay………………………………. 11
3. Proposed Method: FTCP……………………………………………………. 12
3.1 FTCP Concepts and Implementation…………………………………... 12
Unfairness in TCP Congestion Avoidance State……………………… 12
FTCP Congestion Avoidance State…………………………………… 13
Unfairness in TCP Ssthresh Value……………………………………. 15
FTCP Ssthresh Value…………………………………………………. 16
3.2 FTCP Simulation Model……………………………………………….. 19
3.3 FTCP Simulation Results and Discussions…………………………….. 20
3.4 FTCP Conclusions and Future Works…………………………………. 24
4. Proposed Method: Csnoop………………………………………………….. 25
4.1 Csnoop Related Works………………………………………………… 25
4.1.1 Wireless Environment Generic Characteristics………………… 25
Limited Capacity………………………………………………. 25
High Loss Probability and Burst Error Profile………………… 25
Higher End-to-end Delay………………………………………. 26
Frequent Disconnections………………………………………. 26
4.1.2 TCP Problems over Wireless Links…………………………….. 27
Unnecessary Congestion Control……………………………… 27
Inefficiency to Burst Packet Loss……………………………… 27
Longer Delay to Slow Down the Cwnd Recovery Rate……….. 28
4.1.3 Principles and Schemes For TCP over Wireless Links…………. 28
End-to-end Solutions…………………………………………... 30
Fast Retransmit…………………………………………….. 30
Splitting the Connection……………………………………….. 30
I-TCP………………………………………………………. 30
M-TCP……………………………………………………... 31
WTCP……………………………………………………… 32
Link Layer Solutions…………………………………………... 33
ARQ………………………………………………………... 33
FEC………………………………………………………… 33
Snoop………………………………………………………. 34
4.2 Csnoop Concepts………………………………………………………. 37
4.3 Csnoop Implementation………………………………………………... 45
4.4 Csnoop Simulation Model……………………………………………... 47
4.5 Csnoop Simulation Results and Discussions…………………………... 56
4.6 Csnoop Conclusions…………………………………………………… 60
5. Conclusion…………………………………………………………………... 61
6. References…………………………………………………………………... 63

List of Figures Page
Chapter 2 Figure 1 An example of duplicate Acks………………………… 7
Figure 2 TCP cwnd variation of different states………………… 9
Chapter 3 Figure 3 FTCP simulated network………………………………. 19
Figure 4 Throughputs of different protocols in different RTTs…. 21
Figure 5 Fairness index values and bandwidth utilization of different protocols……………………………………... 22
Chapter 4 Figure 6 Flowchart for Snoop_data……………………………... 35
Figure 7 Flowchart for Snoop_ack……………………………… 36
Figure 8 Csnoop queue and elements…………………………… 42
Figure 9a An example of Csnoop calculation……………………. 43
Figure 9b An example of Csnoop calculation……………………. 44
Figure 9c An example of Csnoop calculation……………………. 45
Figure 10 Flowchart for Csnoop_ack…………………………….. 46
Figure 11 Csnoop simulated network…………………………….. 47
Figure 12 A typical bursty losses situation……………………….. 51
Figure 13 Response of TCP Reno (None)………………………... 51
Figure 14 Response of Snoop (Single style)……………………... 52
Figure 15 Response of Csnoop (Continuous style)………………. 53
Figure 16 Response of Allsnoop (All style)……………………… 54
Figure 17 Response of Csnoop-predict (Predictive style)………... 55
Figure 18 Throughputs of different protocols in different mean bad time duration……………………………………… 56
Figure 19 Base station average cached packets size of different protocols………………………………………………... 58



List of Tables Page
Chapter 2 Table 1 Actions of TCP congestion control states……………... 10
Chapter 3 Table 2 TCP cwnd variation of different RTT in congestion avoidance state………………………………………… 13
Table 3 FTCP cwnd variation of different RTT in congestion avoidance state………………………………………… 14
Table 4 TCP cwnd variation of different RTT after timeout state…………………………………………………….. 16
Table 5 FTCP cwnd variation of different RTT after timeout state without modifying ssthresh value………………... 18
Table 6 FTCP cwnd variation of different RTT after timeout state with modifying ssthresh value…………………… 18
Table 7 FTCP simulation results……………………………….. 21
Table 8 Comparison of TCP enhancements over wireless network………………………………………………… 37
Chapter 4 Table 9 Csnoop simulated parameters………………………….. 48
Table 10 Summary of protocols…………………………………. 50

Table 11 Throughputs of different protocols in different mean bad time duration………………………………………. 56
Table 12 Base station average cached packet size of different protocols……………………………………………….. 59

參考文獻 References
[1] J. Postel, “Transmission Control Protocol,” IETF RFC 793, Sept. 1981:available online at http://www.ietf.org/rfc/rfc793.txt
[2] S. Floyd, T. Henderson, “The NewReno Modification to TCP’s Fast Recovery Algorithm,” RFC 2582, Internet Request for Comments 2582, April 1999:available online at http://www.ietf.org/rfc/rfc2582.txt
[3] L.S Brakmo, and L.L. Peterson, “TCP Vegas: End to end congestion avoidance on a global Internet,” IEEE J. Sel. Areas Commun, vol. 13, no. 8, pp.1465-1480, Oct. 1995.
[4] M. Mathis, J. Mahdavi, S. Floyd, A. Romanow, “TCP Selective Acknowledgement Options,” RFC 2018, Internet Request for Comments 2018, October 1996:available online at http://www.ietf.org/rfc/rfc2018.txt
[5] W. Stevens, “TCP Slow Start, Congestion Avoidance, Fast Retransmit, and Fast Recovery Algorithms,” RFC 2001, January 1997:available online at http://www.ietf.org/rfc/rfc2001.txt
[6] V. Javobson, “Congestion Avoidance and Control,” ACM SIGCOMM ’88, pp. 273-288, 1988
[7] F. LEFEVRE, G. VIVIER, “Understanding TCP’s behavior over wireless links,” Communications and Vehicular Technology, 2000, SCVT-200, Symposium on, IEEE 2000, Page(s): 123 –130
[8] R. Caceres, L. Iftnode, “Improving the performance of reliable transport protocols in mobile computing environments,” Selected Areas in Communications, IEEE Journal on, Volume: 13 Issue: 5, June 1995, Page(s): 850 –857
[9] A. Bakre, B.R. Badrinath, “I-TCP: indirect TCP for mobile hosts,” Distributed Computing Systems, 1995. Proceedings of the 15th International Conference on, 1995, Page(s): 136 -143
[10] K. Brown, S. Singth, “M-TCP: TCP for mobile cellular networks,” Computer Communication Review (a publication of ACM SIGCOMM), volume 27(5), October 1997
[11] K. Ratnam and I. Matta. “WTCP: An efficient mechanism fro improving TCP performance over wireless links,” Computers and Communications, 1998. ISCC '98. Proceedings. Third IEEE Symposium on, 1998, Page(s): 74 –78
[12] K. Ratnam and I. Matta. “Effect of local retransmission at wireless access points on the round trip time estimation of TCP,” Simulation Symposium, 1998. Proceedings. 31st Annual, 1998, Page(s): 150 -156
[13] H. Balakrishnan, S. Seshan, R. Katz, “Improving reliable transport and handoff performance in cellular wireless networks,” ACM wireless networks 1, Page(s): 469-481, December 1995.
[14] H. Balakrishnan, V.N. Padmanabhan, S. Seshan, and R.H. Katz. “A comparison of mechanisms for improving TCP performance over wireless links,” Networking, IEEE/ACM Transactions on, Volume: 5 Issue: 6, Dec. 1997, Page(s): 756 -769
[15] K. Kurata, G. Hasegawa, and M. Murata, “Fairness comparisons between TCP Reno and TCP Vegas for future deployment of TCP Vegas,” Proc. INET 2000, June 2000
[16] G. Hasegawa, M. Murata, and Hideo Miyahara, “Fairness and stability of congestion control mechanisms of TCP,” Proc. IEEE INFOCOM ’99, Page(s): 1329-1336, March 1999
[17] G. Hasegawa, M. Murata, “Survey on Fairness Issues in TCP Congestion Control Mechanisms,” IEICE TRANS. COMMUN. vol. E84-B, NO.6, Page(s):1461-1471, June 2001
[18] UCB/LBNL/VINT Network Simulator (ns-2) available on-line at http://www.isi.edu/nsnam/ns
[19] R. Jain, “The Art of Computer Systems Performance Analysis: Techniques for Experimental Design, Measurement, Simulation, and Modeling,” New York: Wiley, 1991.
[20] D. Huang, J.J. Shi, “Performance of TCP over radio link with adaptive channel coding and ARQ,” Vehicular Technology Conference, 1999 IEEE 49th, Volume: 3, 1999, Page(s): 2084 -2088 vol.3
[21] M. Allen, V. Paxon, W. Stevens, “TCP Congestion Control,” RFC 2581, April 1999:available online at http://www.ietf.org/rfc/rfc2581.txt
電子全文 Fulltext
本電子全文僅授權使用者為學術研究之目的,進行個人非營利性質之檢索、閱讀、列印。請遵守中華民國著作權法之相關規定,切勿任意重製、散佈、改作、轉貼、播送,以免觸法。
論文使用權限 Thesis access permission:校內校外完全公開 unrestricted
開放時間 Available:
校內 Campus: 已公開 available
校外 Off-campus: 已公開 available


紙本論文 Printed copies
紙本論文的公開資訊在102學年度以後相對較為完整。如果需要查詢101學年度以前的紙本論文公開資訊,請聯繫圖資處紙本論文服務櫃台。如有不便之處敬請見諒。
開放時間 available 已公開 available

QR Code