Rate-Adaptive TCP Spoofing with Segment Aggregation over Asymmetric Long Delay Links

Authors:
Jihyun Lee;Hyungyu Park;Kyungshik Lim
Affiliations:
School of EECS, Kyungpook National University, Daegu, Korea 702-701;School of EECS, Kyungpook National University, Daegu, Korea 702-701;School of EECS, Kyungpook National University, Daegu, Korea 702-701
Venue:
Information Networking. Towards Ubiquitous Networking and Services
Year:
2008

Citing 4
Cited 0

Improving TCP/IP performance over wireless networks

MobiCom '95 Proceedings of the 1st annual international conference on Mobile computing and networking
The effects of asymmetry on TCP performance

Mobile Networks and Applications
Window-Based Error Recovery and Flow Control with a Slow Acknowledgement Channel: A Study of TCP/IP Performance

INFOCOM '97 Proceedings of the INFOCOM '97. Sixteenth Annual Joint Conference of the IEEE Computer and Communications Societies. Driving the Information Revolution
On TCP performance in a heterogeneous network: a survey

IEEE Communications Magazine

Quantified Score

Hi-index	0.00

Visualization

Abstract

To improve the TCP performance over long delay links, TCP spoofing has been mainly proposed as a particular solution. Even if it can reduce the duration of the slow-start, it provides the limited maximum throughput of approximately 1 Mbps with the maximum window size of 64KB. In this paper, we propose a rate control algorithm, called Rate-Adaptive spoofing (RA-spoofing) that adjusts virtual window dynamically depending on the number of other TCP connections sharing the link, thereby fully utilizing the available bandwidth and maximizing throughput rates. Tightly coupled with RA-spoofing, we also propose a Segment Aggregation (SA) mechanism to overcome the bandwidth asymmetry problem. SA is a multiplexing process that a number of TCP segments are transformed into the corresponding number of IP fragments directly, based on segment aggregation factor , which results in one true ACK from the destination. The results of the performance analysis provide a good evidence to demonstrate the efficiency of our mechanisms under different assumptions of bandwidth asymmetric ratio and error rate.