MaxNet: Faster Flow Control Convergence

MaxNet is a distributed congestion control architecture in which only the most severely bottlenecked link on the end-to-end path generates the congestion signal that controls the source rate. This is unlike SumNet networks, such as the current Internet or REM, where all of the bottlenecked links on the end-to-end path add to the congestion signal (by packet marking or dropping). Previously, we have shown that MaxNet results in MaxMin like rate allocation and is stable for arbitrarily large networks. In this paper we analyze the small-signal convergence speed of MaxNet. We show that MaxNet is able to converge faster than the SumNet architecture. Faster convergence results in less delay jitter, higher utilisation and lower buffer sizes. Furthermore, we show that MaxNet decouples the control, so that each pole position depends only on parameters of one bottleneck link and of the sources controlled by that bottleneck, enabling optimal pole placement.

• Type: Conference
• Year: 2004
• Where: NETWORKING
• Authors: Bartek P. Wydrowski, Lachlan L. H. Andrew, Iven M. Y. Mareels
• Keywords: Congestion Control Architecture, Congestion Signal, End-to-end Path, NETWORKING 2004, NETWORKING 2007