Towards understanding residential internet traffic: From packets to services

Abstract

Traffic characterization is a key aspect for Internet Service Providers (ISPs) to tune up their networks according to their customers' requirements. Despite the significant gap between business and residential customers, a large body of literature measurements is performed at higher observation points and only few studies have examined residential traffic characteristics. Furthermore, traffic analysis granularity is often too coarse to tackle observed growth of applications and services. In this paper, we present a fine-grained analysis of a real residential traffic dataset collected in France and provided by a major ISP involving more than 40,000 customers. Our findings provide useful insights of usage patterns, applications distribution and traffic characteristics. Among these findings, HTTP and SSL traffic accounts for more than 73% of total data volume. Also, YouTube generates almost 70% of real time entertainment flows.