Large Scale Benchmarking of Broadband Access Networks: Issues, Methodologies, and Solutions

Abstract

The Internet today counts about 2.1 billion users worldwide, thus representing a highly complex system on which huge amounts of data are transferred every day over the IP protocol. Most people rely on Internet connectivity for everyday activities, making broadband access an essential resource, whose performance have not been widely studied in literature. On one hand, evaluating IP networks performance is a challenging task which is made even more difficult by the heterogeneity of underlying protocols and technologies. On the other hand, the complexity of emerging applications makes it difficult to understand the relation between network and user-perceived performance. In this thesis we address the evaluation of IP networks performance with a specific focus on broadband access networks. We start presenting an overview of the Internet scenario with respect to the heterogeneity of access technologies and the complexity of new generation applications. Hence, we conduct research activities on both the analysis and characterization of traffic generated by new-generation applications and the identification of relevant metrics, methodologies, techniques and tools to evaluate network performance. We conduct a study on the evaluation of broadband access networks performance, in which we propose a taxonomy of existing approaches and define the guidelines to build an architecture with ideal characteristics to measure access networks performance on a large scale. As a result of the preceding analysis we identified the ideal characteristics for an architecture able to evaluate the performance of broadband access networks on a large scale. Following such guidelines, we designed and implemented two architectures respectively adopting the router- and client-based approaches, in order to evaluate their potentialities and the eventual advantages deriving from a combined approach. Thanks to this activity we devised new methodologies and techniques to cope with several challenges arose during the different phases from the design to the deployment. Thanks to the data collected through these architectures, we also found some interesting results about real access networks, by identifying and characterizing relevant issues affecting most of them.