A Machine Learning Approach to Estimating Queuing Delay on Routers: The Multiple-Hop Case

Compression and Decompression (CoDe) is an active, state-of-the-art scheme for measuring queuing delay on Internet routers that utilizes an unsupervised learning algorithm for data processing. In this paper, we present a comparative evaluation of CoDe over a multiple-hop path under demanding traffic conditions using ns-3 simulation. This is the first evaluation of this kind because prior study of the scheme only considered a single-hop path, reflecting a limited performance evaluation. Our simulation data shows that CoDe measures queuing delay on intermediate routers along a multiple-hop path with high accuracy. It also outlines some measurement challenges prevalent in other existing schemes.