Home // ICSNC 2014, The Ninth International Conference on Systems and Networks Communications // View article

Inferring TCP Congestion Control Algorithms by Correlating Congestion Window Sizes and their Differences

Authors:
Toshihiko Kato
Atsushi Oda
Shun Ayukawa
Celimuge Wu
Satoshi Ohzahata

Keywords: TCP congestion control algorithms; passive monitoring; congestion window

Abstract:
Recently, according to the diversification of network environments, a lot of TCP congestion control mechanisms have been introduced. They define how a TCP sender increases the congestion window (cwnd) during no congestion and decreases cwnd when it detects congestion. Since the congestion control algorithms affect the performance of the Internet, it is important to know which algorithms are used widely. This paper proposes a scheme to infer the algorithm for individual TCP flows by examining the packet trace passively captured for the flows. Our scheme estimates cwnd values in RTT intervals and correlates the cwnd values and the differences of consecutive cwnd values. Our scheme aims to infer many of recently proposed congestion control algorithms, which have been out of scope in the conventional passive approaches. Our scheme adopts a simple approach just correlating cwnds and their differences, in contrast with the conventional approaches which estimate the internal TCP behaviors based on packet traces. This paper describes the details of our scheme and shows the results where our scheme is applied to an iPhone TCP communication.

Pages: 42 to 47

Copyright: Copyright (c) IARIA, 2014

Publication date: October 12, 2014

Published in: conference

ISSN: 2163-9027

ISBN: 978-1-61208-368-1

Location: Nice, France

Dates: from October 12, 2014 to October 16, 2014