Faculty Advisor or Committee Member

Craig E. Wills, Advisor

Faculty Advisor or Committee Member

Mark L. Claypool, Reader




With the rising popularity of Web-based applications, the Web browser platform is becoming the dominant environment in which users interact with Internet content. We investigate methods of discovering information about network performance characteristics through the use of the Web browser, requiring only minimal user participation (navigating to a Web page). We focus on the analysis of explicit and implicit network operations performed by the browser (JavaScript XMLHTTPRequest and HTML DOM object loading) as well as by the Flash plug-in to evaluate network performance characteristics of a connecting client. We analyze the results of a performance study, focusing on the relative differences and similarities between download, upload and round-trip time results obtained in different browsers. We evaluate the accuracy of browser events indicating incoming data, comparing their timing to information obtained from the network layer. We also discuss alternative applications of the developed techniques, including measuring packet reception variability in a simulated streaming protocol. Our results confirm that browser-based measurements closely correspond to those obtained using standard tools in most scenarios. Our analysis of implicit communication mechanisms suggests that it is possible to make enhancements to existing “speedtest” services by allowing them to reliably determine download throughput and round-trip time to arbitrary Internet hosts. We conclude that browser-based measurement using techniques developed in this work can be an important component of network performance studies.


Worcester Polytechnic Institute

Degree Name



Computer Science

Project Type


Date Accepted





network measurement, Web browsers, Browsers (Computer programs)