Mobile Web Page Performance: Bottlenecks, Visualization, and QoE

Duke Computer Science Colloquium
Speaker Name
Aruna Balasubramanian
Date and Time
D106 LSRC, Duke
Pizza lunch will be served at 11:45.

Page load performance is crucial to the Internet ecosystem since a significant fraction of Internet content is consumed as Web pages. However, Web page loads are extremely slow, especially on mobile devices. While many optimizations for improving page load performance exists, there is no silver bullet. Often, an optimization that improves performance of one page hurts performance of another. The key problem is that the different Web pages have different bottlenecks, and optimizing pages without understanding their bottlenecks is futile. In this talk, I will describe WProf, a tool that identifies the bottleneck and critical path during the page load process using dependency analysis. Using WProf, we show that the main bottleneck when loading pages on a mobile browser is computation, whereas network activities are the main bottleneck when loading pages on desktop browsers. We have recently developed a visualization tool to easily analyze these bottlenecks (the tool is available at

I will next describe our work Webgaze that uses eye gaze tracking to improve Web Quality of Experience (QoE). We show that traditional Web performance metrics have little correlation with user experience. Yet, Web optimizations are designed to improve these traditional metrics. I will describe WebGaze, a practical system that leverages eye tracking to improve Web QoE. More details are at

Short Biography

Aruna Balasubramanian is an Assistant Professor at Stony Brook University. She received her Ph.D from the University of Massachusetts Amherst, where her dissertation won the UMass outstanding dissertation award and was the Sigcomm dissertation award runner up. She works in the area of networked systems. Her current work consists of three threads: (1) significantly improving mobile Web performance and (2) improving the usability and privacy of mobile systems, and (3) designing platforms to support deep learning on mobile and embedded devices. She is the recipient of a Ubicomp best paper award, a Computing Innovation Fellowship, a Microsoft Graduate Research Fellowship, a Google research award, and the Applied Networking Research Prize. She is passionate about improving the diversity in Computer Science and runs a Girls Who Code Club, founded the WPhD group for Women PhD students at Stony Brook, and is an active member of the N2Women group.

Landon Cox