[en] We study the problem of predicting Internet path changes and path performance using traceroute and machine-learning techniques. Path changes are frequently linked to path inflation and performance degradation. Therefore, predicting their occurrence could improve the analysis of path dynamics using traceroute. By relying on neural networks and using empirical distribution based input features, we show that we are able to predict (i) the remaining life time of a path before it actually changes, and (ii) the number of path changes in a certain time slot with relatively high accuracy. We also show that it is possible to predict path performance in terms of latency, opening the door to novel, machine-learning-based approaches for RTT prediction.
Disciplines :
Sciences informatiques
Auteur, co-auteur :
Wassermann, Sarah ; Université de Liège - ULiège > Master sc. informatiques, à fin.
Casas, Pedro
Donnet, Benoît ; Université de Liège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Algorithmique des grands systèmes
Langue du document :
Anglais
Titre :
Machine Learning based Prediction of Internet Path Dynamics
Date de publication/diffusion :
décembre 2016
Nom de la manifestation :
ACM CoNEXT Student Workshop
Lieu de la manifestation :
Irvine, Etats-Unis - Californie
Date de la manifestation :
12 décembre 2016
Manifestation à portée :
International
Titre de l'ouvrage principal :
ACM CoNEXT Student Workshop: Irvine 12 décembre 2016