direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

Fabian Schneider's Publications

Analysis of New Trends in the Web from a Network Perspective
Citation key S-ANTWNP-09
Author Schneider, Fabian
Year 2010
School Technische Universität Berlin, Berlin, Germany
Abstract Over the last five years, several trends have changed the landscape of the World Wide Web, forming the new ``Web 2.0''. The advent of user generated content (blogs and wikis), the popularity of multimedia (e.g., YouTube and MySpace), and the penetration of Google¿s services (maps, mail, etc.) are commonly noticeable. In particular, the recent popularity of Online Social Networks (OSNs, e.g., Facebook and LinkedIn) has caused a fundamental change in how the Internet is used. For example, certain OSN users are only using the OSN internal messaging instead of email. This motivates us to examine the usage of these new Web trends and determine their impact on the network. First, we present a traffic study of several Web 2.0 applications including Google Maps, modern Web-based email, and social networking Websites, and compare their traffic characteristics with the ambient HTTP traffic. We highlight the key differences between Web 2.0 traffic and all HTTP traffic. As such, our work elucidates the changing face of one of the most popular applications on the Internet: The World Wide Web. We find that ``Web 2.0'' applications unleash new HTTP traffic patterns which differ from the conventional HTTP request-response model. In particular, asynchronous pre-fetching of data in order to provide a smooth Web browsing experience and richer HTTP payloads (e.g., JavaScript libraries) of Web 2.0 applications induce larger, heavier, and more bursty traffic on the underlying networks. Next, we focus on Online Social Networks. OSNs have already attracted more than half a billion users. However, our understanding of which OSN features attract and keep the attention of these users is poor. Studies thus far have relied on surveys or interviews of OSN users or focused on static properties, e.g., the friendship graph, gathered via sampled crawls. In this thesis, we study how users actually interact with OSNs by extracting anonymized clickstreams from passively monitored network traffic. Our characterization of user interactions within the OSN for four different OSNs (Facebook, LinkedIn, Hi5, and StudiVZ) focuses on feature popularity, session characteristics, and the dynamics within OSN sessions. We find, for example, that users commonly spend more than half an hour interacting with the OSN. Yet, the byte contributions per OSN session are relatively small. Subsequently, we look into mobile hand-held device (MHD) usage that is observed when such devices are used at home. Our characterization of the traffic shows that mobile Apple devices (i.e., iPhones and iPods) are, by a huge margin, the most commonly used MHDs and account for most of the traffic. We find that MHD traffic is dominated by multimedia content and downloads of mobile applications. Finally, inspired by the finding that Network News Transport Protocol (NNTP) traffic is responsible for up to 5\% of residential network traffic we investigate today¿s Usenet usage. We find that NNTP is intensively used by a small fraction of the residential broadband lines that we study and that almost all traffic is originated by NNTP servers that require a monthly fee subscription. The accessed content resembles what one might expect from file-sharing systems–archives and multimedia files. Accordingly, it appears that NNTP is used by some as a high performance alternative to traditional P2P file-sharing options such as eDonkey or BitTorrent. The analyses of this thesis are based on anonymized packet level recordings of real Internet traffic collected at different vantage points and from different user populations. From these recordings we extract traces of protocol events and activities (e. g., new TCP connections or application layer requests and responses). Next, all traffic related to a new trend is identified and then compared to the overall traffic. This allows us to predict the impact on the traffic in case these trends increase or decrease in popularity. Packet capture in high speed networks is challenging. Limitations (e.g., bus or memory bandwidth, OS capturing stack) prevent comprehensive packet capture in these environments. The endeavor to perform packet capture with commodity hardware requires us to identify and then overcome some of the performance limitations. Knowing the limitations, we propose several possibilities to split or reduce the analysis load.
Bibtex Type of Publication Doktorarbeit
Link to publication Link to original publication Download Bibtex entry

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions