DOI QR코드

DOI QR Code

An Effective Solution to Overcome the Restriction of SACK Blocks' Number in TCP SACK

오프셋을 활용한 효율적인 TCP SACK 메커니즘

  • 최린 (경북대학교 컴퓨터과학과) ;
  • 홍충선 (경희대학교 컴퓨터공학과)
  • Published : 2005.12.01

Abstract

TCP SACK is the unique mechanism to reflect the situation of sink's sequence space, some TCP variants and proposals can perform in conjunction with SACK mechanism for achieving optimal performance. By definition of RFC 2018, however, each contiguous block of data queued at the data receiver is defined in the SACK option by two 32-bit unsigned integers in network byte order. Since TCP Options field has a 40-byte maximum length, when error bursts now, we note that the available option space may not be sufficient to report all blocks present in the receiver's queue and lead to unnecessarily force the TCP sender to retransmit Packets that have actually been received by TCP sink. For overcoming this restriction, in this thesis, a new solution named 'one-byte offset based SACK mechanism' is designed to further improve the performance or TCP SACK and prevent those unwanted retransmissions. The results or both theory analysis and simulation also show that his proposed scheme operates simply and more effectively than the other existing schemes by means of the least bytes and most robust mechanism to the high packet error rates often seen in networks environment.

TCP SACK은 sink의 순차적인 필드 상태를 나타내는 유일한 메커니즘이며, 여러 가지 변형된 TCP들은 최적의 성능을 위해서 SACK 메커니즘을 적용할 수 있다. RFC 2018에서 SACK 옵션은 수신자 측에 쌓여진 데이터 큐 각각의 연속된 블록으로 2개의 32비트로 정의되어 있다. TCP 옵션 필드는 최대 40바이트 길이를 가지기 때문에 에러가 발생하였을 때, TCP 수신자 큐에 있는 모든 데이터 블록들을 알려줄 수 있는 사용 가능한 옵션 공간이 충분하지 않으며, TCP 송신자가 TCP sink에 의해서 수신된 패킷들을 불필요하게 재 전송하게 된다. 이러한 문제들을 해결하기 위해서 본 논문에서는 TCP SACK의 성능을 향상시키고 불필요한 재전송을 제거하기 위해서 "one-byte offset based SACK mechanism" 이라는 새로운 방식을 제시한다. 제안된 방식의 분석과 시뮬레이션 결과 제안된 방식은 최소한의 바이트를 사용하기 때문에 다른 메커니즘들보다 오버헤드를 줄였고, 유무선 통합 환경에서 에러율이 적은 효율적인 메커니즘임을 입증하였다.

Keywords

References

  1. M. Mathis, J. Mahdavi, S. Floyd, A Romanow, 'TCP selective acknowledgment and options', RFC 2018, IETF, October, 1996
  2. Balan, R.K.; Lee, B.P.; Kumar, K.R.R.; Jacob, L.; Seah, W.K.G.; Ananda, A.L. 'TCP HACK: TCP Header Checksum Option to Improve Performance over Lossy Links', lNFOCOM 2001. IEEE, Volume: 1, 22-26 April, 2001 Pages : 309-318 Vol.1 https://doi.org/10.1109/INFCOM.2001.916713
  3. V. Jacobson, R. Braden, D. Borman, 'TCP Extensions for High Performance', RFC 1323, IETF, May, 1992
  4. S. Floyd, J. Mahdavi, M. Mathis, M. Podolsky, 'An Extension to the Selective Acknowledgement(SACK) Option for TCP', RFC 2883, IETF, July, 2000
  5. M. Allman, NASA Glenn, V. Paxson, W. Stevens 'TCP Congestion Control', RFC 2581, IETF, April 1999
  6. Jacobson, V., 'Congestion Avoidance and Control', Computer Communication Review, Vol.18, No.4, pp.314-329, Aug., 1988 https://doi.org/10.1145/52325.52356
  7. K. Fall, S. Floyd, 'Simulation based Comparisons of Tahoe, Reno, and SACK TCP', Computer Communications Review, Vol.26, pp.5-21, 1996 https://doi.org/10.1145/235160.235162
  8. Floyd, S. and T. Henderson, 'The NewReno Modification to TCP's Fast Recovery Algorithm', RFC 2582, April, 1999
  9. Braden, R., 'Requirements for Internet Hosts-Communication Layers', STD 3, RFC 1122, October, 1989
  10. Luigi A. Grieco and Saverio Mascolo, 'Performance Evaluation and Comparison of Westwood+, New Reno, and Vegas TCP Congestion Control', ACM, 2004 https://doi.org/10.1145/997150.997155
  11. S. Mascolo, C. Casetti, M. Gerla, M. Sanadidi, and R. Wang, 'TCP Westwood: End-to-end bandwidth estimation for efficient transport over wired and wireless networks,' in ACM Mobicom 2001, Rome, Italy, July, 2001
  12. Matthew Mathis, Jamshid Mahdavi, Sally Floyd, and Allyn Romanow. 'TCP Selective Acknowledgment Options,'. (Internet draft, work in progress), 1996
  13. K.N. Srijith, Lillykutty Jacob, A.L. Ananda, 'Worst-Case Performance Limitation of TCP SACK and a Feasible Solution', IEEE Journal on Selected Areas in Communications, 2002
  14. Network Simulator(ns-2), available from http://www.isi.edu/nsnam/ns/