A Unified Approach to Internet Performance Measurement
The goal of this research is to develop novel, reliable, and scalable QoE crowdtesting methodologies to diagnose QoE degradation issues in video streaming and video conferencing applications. We propose to develop and apply a fundamentally new approach to Internet performance measurement research that will narrow critical gaps between QoS and QoE measurements.
Principal Investigator: Ka Pui Mok
Funding source: CNS-2133452 Period of performance: December 15, 2021 - November 30, 2024.
The use of online applications, such as video streaming and video conferencing, significantly surged during the COVID-19 pandemic to facilitate virtual meetings and classroom instruction. Service providers and regulators face increasing pressure to understand the quality of experience (QoE) of these applications. Poor QoE can greatly degrade the usability of applications and thus impair the effectiveness of communication and learning. But scientific measurement of QoE faces a daunting set of challenges.
We formulate two tasks to tackle the challenges.
The first task is to design and implement new experiments to extend the capability of our gamified web platform – Quality of Internet Consumer Experience (QUINCE) – to crowdsource network measurements and QoE assessments from users of major video streaming platforms.
The second task is to analyze the correlation between network performance metrics and reported QoE. In addition to network measurements embedded in QUINCE, we will leverage external measurement platforms capable of inferring evidence of interdomain congestion and reactively triggering throughput measurements to speed test servers from the cloud. The combined data will support our analysis of the dynamics of video streaming/conferencing QoE when congestion occurs in interconnections and access networks.
212Task 1: Extend the capability of QUINCE
|1.1||Embed videos with player APIs|
|1.2||Implement video conferencing QoE test|
|1.3||Enhance gamification elements|
|1.4||Conduct pilot studies and refine the system|
353Task 2: Diagnose QoE degradation
|2.1||Deploy crowdsourcing experiments|
|2.2||Perform QoE degradation analysis|
|2.3||Carry out evaluation and validation|
Acknowledgment of awarding agency’s support
This material is based on research sponsored by the National Science Foundation (NSF) grant CNS-2133452. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of NSF.