A new end-to-end probing and analysis method for estimating bandwidth bottlenecks

TLDR

The paper introduces TOPP, a network‑friendly bandwidth measurement method that uses active probing and segmented regression analysis. TOPP estimates two complementary available bandwidth metrics and the congested link bandwidth, and can compute bandwidth estimates for multiple congested hops from a single end‑to‑end probe session. Unlike traditional packet‑pair methods, TOPP’s estimates are not constrained by probe injection rate and it can detect bottlenecks that other techniques, such as C‑probe, miss.

Abstract

We present a network friendly bandwidth measurement method, TOPP, that is based on active probing and includes analysis by segmented regression. This method can estimate two complementing available bandwidth metrics in addition to the link bandwidth of the congested link. Contrary to traditional packet pair estimates of the bottleneck link bandwidth, our estimate is not limited by the rate at which we can inject probe packets into the network. We also show that our method is able to detect bottlenecks that are invisible to methods such as the C-probe. Further more, we describe scenarios where our analysis method is able to calculate bandwidth estimates for several congested hops based on a single end-to-end probe session.