Characterizing broadband user behavior

2011

The campus network is a microcosm of the entire Internet, containing various types of users utilizing a variety of Internet applications. This paper characterizes the campus Internet workload for a mid-size university consisting of over 10,000 users. The paper characterizes bandwidth usage, top bandwidth consuming applications and users’ Internet usage patterns on our campus network. The total amount of data transferred between campus and the Internet for an entire year, as well as the data amounts for individual application classes are presented. The paper also distinguishes the significant differences between students and faculty usage patterns. Over the past few years, there has been a major shift in application usage on campus networks. Real-time and bandwidth intensive applications dominate the workload of academic networks. The workload characterization in this paper identifies the degree of impact that these applications have on system resources and user workloads.

GLOBECOM'01. IEEE Global Telecommunications Conference (Cat. No.01CH37270), 2001

In this work the authors show how the behaviour of Web users strongly affects the dynamics of TCP connections in Internet. Analysing actual and systematically generated HTTP traces, it is proved that the time between the download of two pages is critical to determine the re-utilisation of TCP connections. On the other hand, the study also shows that the utilisation of 1.1 version of the HTTP standard does not significantly affect the traffic generated by HTTP 1.0. In this sense, the heavy-tailed nature of the size of HTTP connections can be considered as an invariant property.

The global Internet has seen tremendous growth in terms of nodes and user base as well as of types of applications. One of the most important consequences of this growth is related to an increased complexity of the traffic models experienced in the networks. Each application has a set of unique characteristics in terms of the way it performs its transactions as well as the way transaction processing profile maps onto unique network resource requirements. In order to support Internet applications effectively, it is therefore important to understand and to characterize the application level transactions and also to investigate their scaling properties. Recent advances in high resolution traffic monitoring and analyzing capabilities have enabled us to build up realistic models for the TCP/IP protocol stack with diverse network applications. In this paper we report investigations of classical applications such as FTP, SMTP, and HTTP to evaluate end-to-end performance requirements and ac...

This dissertation deals with monitoring, collecting, analyzing, and modeling of World Wide Web (WWW) traffic and client interactions. The rapid growth of WWW usage has not been accompanied by an overall understanding of models of information resources and their deployment strategies. Consequently, the current Web architecture often faces performance and reliability problems. Scalability, latency, bandwidth, and disconnected operations are some of the important issues that should be considered when attempting to adjust for the growth in Web usage. The WWW Consortium launched an effort to design a new protocol that will be able to support future demands. Before doing that, however, we need to characterize current users' interactions with the WWW and understand how it is being used.

Proceedings of the 33rd Annual Hawaii International Conference on System Sciences, 2000

The World Wide Web (WWW) gives a quick and easy access to an enormous variety of information in remote locations. Sometimes, obtaining the Web pages implies a long time for their retrieval. One of the problems appears because, at the same time, the same copy can be claimed by other users, producing that identical copies of many documents pass through the same network links, whereby the network administrators see a growing utilization that requires bandwidth upgrades, and Web site administrators see a growing server utilization that requires upgrading or replacing servers. The key performance factors to consider are how to reduce the volume of network traffic produced by Web clients and servers, and how to improve the mean response time for WWW users. Mechanisms such as mirroring and caching have been proposed to rescue the Internet by reducing the page waiting time and WWW traffic. Mirroring involves cooperation between the page owners and the mirror sites and thus requires prior arrangements. Caching can be more generally applied where a cache server can be set up to provide closer-to-user services for users who wish to reduce the page access time by selecting the cache servers as their proxy server. This shows that the aim of improving Web performance depends on a deep knowledgement of WWW workload characterization.

Web Information Systems Engineering and Internet Technologies Book Series, 2005

] conducted a comprehensive workload characterization study of Internet Web servers. By analyzing access logs from 6 Web sites (3 academic, 2 research, and 1 industrial) in 1994 and 1995, the authors identified 10 invariants: workload characteristics common to all the sites that are likely to persist over time. In this present work, we revisit the 1996 work by Arlitt and Williamson, repeating many of the same analyses on new data sets collected in 2004. In particular, we study access logs from the same 3 academic sites used in the 1996 paper. Despite a 30-fold increase in overall traffic volume from 1994 to 2004, our main conclusion is that there are no dramatic changes in Web server workload characteristics in the last 10 years. Although there have been many changes in Web technologies (e.g., new protocols, scripting languages, caching infrastructures), most of the 1996 invariants still hold true today. We postulate that these invariants will continue to hold in the future, because they represent fundamental characteristics of how humans organize, store, and access information on the Web.

Smart devices bring us the ubiquitous mobile accessing to Internet, making mobile Internet grow rapidly. Using the mobile traffic data collected at core metropolitan 2G and 3G networks of China over a week, this paper studies the mobile user behavior from three aspects: 1) data usage; 2) mobility pattern; and 3) application usage. We classify mobile users into different groups to study the resource consumption in mobile Internet. We observe that traffic heavy users and high mobility users tend to consume massive data and radio resources simultaneously. Both the data usage and the mobility pattern are closely related to the application access behavior of the users. Users can be clustered through their application usage behavior, and application categories can be identified by the ways to attract the users. Our analysis provides an comprehensive understanding of user behavior in mobile Internet, which may be used by network operators to design appropriate mechanisms in resource provision and mobility management for resource consumers based on different categories of applications. INDEX TERMS Mobile Internet, network traffic, data usage, mobility pattern, user behavior.

Lecture Notes in Computer Science, 2012

We compare Web traffic characteristics of mobile-versus fixed-access end-hosts, where herein the term "mobile" refers to access via cell towers, using for example the 3G/UMTS standard, and the term "fixed" includes Wi-Fi access. It is well-known that connection speeds are in general slower over mobile-access networks, and also that often there is higher packet loss. We were curious whether this leads mobile-access users to have smaller connections. We examined the distribution of the number of bytesper-connection, and packet loss from a sampling of logs from servers of Akamai Technologies. We obtained 149 million connections, across 57 countries. The mean bytes-per-connection was typically larger for fixed-access: for two-thirds of the countries, it was at least onethird larger. Regarding distributions, we found that the difference between the bytes-per-connection for mobileversus fixed-access, as well as the packet loss, was statistically significant for each of the countries; however the visual difference in plots is typically small. For some countries, mobile-access had the larger connections. As expected, mobile-access often had higher loss than fixed-access, but the reverse pertained for some countries. Typically packet loss increased during the busy period of the day, when mobile-access had a larger increase. Comparing our results from 2010 to those from 2009 of the same time period, we found that connections have become a bit smaller.

Proceedings Fourth IEEE International Workshop on Advanced Issues of E-Commerce and Web-Based Information Systems (WECWIS 2002)

The World Wide Web has achieved immense popularity in the business world. It is thus essential to characterize the traffic behavior at these sites, a study that will facilitate the design and development of high-performance, reliable e-commerce servers. This paper makes an effort in this direction. Aggregated traffic arriving at a Businessto-Business (B2B) and a Business-to-Consumer (B2C) e-commerce site was collected and analyzed. High degree of self-similarity was found in the traffic (higher than that observed in general Web-environment). Heavy-tailed behavior of transfer times was established at both the sites. Traditionally this behavior has been attributed to the distribution of transfer sizes, which was not the case in B2C space. This implies that the heavy-tailed transfer times are actually caused by the behavior of back-end service time. In B2B space, transfer-sizes were found to be heavy-tailed. A detailed study of the traffic and load at the back-end servers was also conducted and the inferences are included in this paper.

11th IEEE/ACM International Symposium on Modeling, Analysis and Simulation of Computer Telecommunications Systems, 2003. MASCOTS 2003., 2003