End-to-End Bandwidth Estimation for TCP Congestion Control in Packet Networks

Today TCP/IP congestion control implements the additive increase/multiplicative decrease (AIMD) paradigm to probe network capacity and obtain a “rough” but robust measurement of the best effort available bandwidth. To improve the efficiency of congestion control, Westwood TCP proposes an additive increase/adaptive decrease paradigm that adaptively sets the transmission rate at the end of the probing phase in order to exactly match the bandwidth used at the time of congestion, which is the definition of best-effort available bandwidth in a connectionless packet network. This work addresses the challenging issue of estimating the best-effort bandwidth available for a TCP/IP connection by properly counting and filtering the stream of acknowledgments packets using discrete-time filters. The estimate has to provide the low frequency components of the used bandwidth because congestion does not depend on high frequency components. Moreover, the filtering algorithm has to deal with non-uniform sampling intervals due to the fact that the stream of ACK packets is asynchronous. In particular, ACK packets experience congestion along the backward path and reach the sender bunched in groups, which is the ACK compression phenomenon. The main resul t shown in this work is that, in order to implement a low-pass filter in packet networks, it is necessary to implement an anti ACK compression algorithm, which plays the role of a classic anti-aliasing filter. Moreover, a comparison of time-invariant and time-varying discrete filters to be used after the anti-aliasing algorithm is developed. Finally, simulation results show the effectiveness of the proposed bandwidth estimation algorithms, which enhance the efficiency of TCP congestion control especially over lossy links.