Skip to Main content Skip to Navigation

Application Level Performance in Wired and Wireless Environments

Abstract : The interest in traffic measurement and analysis has increased tremendously and provides us with new ways to understand, operate and improve network performance. The heterogeneity of the Internet is constantly increasing, with new access technologies, new client devices and with more and more services and applications. On the other hand, the interest of the research community to measure enterprise network performance has grown, due to a complexity that sometimes rivals Internet. These subjects, of crucial importance for service providers, network managers and companies have already received substantial attention in the research community. Despite these efforts, a number of issues still remain unsolved. This thesis is concerned with TCP traffic, which carries the large majority of the Internet's traffic. While analyzing the performance of TCP transfers, we focused on the connections that correspond to valid and complete transfers, from the TCP perspective. The present work consists of three parts dealing with various aspects of the challenging task of, revisiting TCP performance, performance study and anomaly detection. In the first part, we revisit most important works and discuss problems faced when we studied TCP performance. We present an overview of the impact of the application, on the TCP transfers. We show that while losses can have a detrimental impact on short TCP transfers, the application significantly affects the transfer time of almost all short and even long flows. In this part we show that measurements from passively collected traces can be biased by specific technologies implemented in Cellular networks to boost performance and control users activity. In the second part, we compare the performance of Cellular, FTTH and ADSL accesses with traces collected on access networks under the control of the same ISP. We shows that a study of classical performance parameters does not lead to a full understanding of client perceived throughput. Then, we propose and validate a method that drills down into the data transfer of each well-behaved connection. The Data time break-down approach automatically extracts the application, access, server and client behavior impacts from passively observed TCP transfers. It also groups together, with an appropriate clustering algorithm, the transfers that have experienced similar performance over different access technologies. We then characterize some salient aspects of analyzing enterprise traffic and we provide an overview of problems. In the last part, we focus on the issue of profiling anomalous TCP connections that are defined as functionally correct TCP connections but with abnormal performance. Our method enables to pinpoint the root cause of the performance problem, which can be either losses or some idle times during data preparation or transfer. We apply this methodology to several traces corresponding to Internet and enterprise traffic. We demonstrate the existence of specific strategies to recover from losses on Cellular networks that seem more efficient than what is done currently in wired networks.
Complete list of metadatas

Cited literature [98 references]  Display  Hide  Download
Contributor : Aymen Hafsaoui <>
Submitted on : Tuesday, February 14, 2012 - 11:40:07 AM
Last modification on : Friday, July 31, 2020 - 10:44:07 AM
Long-term archiving on: : Wednesday, December 14, 2016 - 5:01:06 AM


  • HAL Id : pastel-00669973, version 1


Aymen Hafsaoui. Application Level Performance in Wired and Wireless Environments. Networking and Internet Architecture [cs.NI]. Télécom ParisTech, 2011. English. ⟨pastel-00669973⟩



Record views


Files downloads