VFAST TCP: An improvement of FAST TCP

Abstract

We describe a delay-based end-to-end congestion control algorithm, called Very FAST TCP (VFAST), which is an extension of FAST TCP. The idea behind this extension is to smoothly estimate the round-trip-time, which eliminates throughput and queue oscillation when round-triptime fluctuates. In this context, we introduce a simulation and an evaluation of the suggested scheme, by comparing our VFAST prototype with FAST in terms of throughput, fairness, stability, RTT and adaptivity to changes in network. The preliminary results are promising.