Quality of experience (QoE)

Games QoE

Quality of Experience (QoE) assessment for video games is known for being a heavy-weight process, typically requiring the active involvement of several human players. To disrupt the status quo, we propose to remove human players from the loop and instead exploit Deep Reinforcement Learning (DRL) agents to play games under varying network conditions, using achievable scores under bad network conditions a proxy of user QoE.

To know more:

Web QoE

We introduce methods to accurately compute any Web Quality of Experience (QoE) metrics from encrypted traffic using artificial intelligence techniques, as well as to

to efficiently compute approximate a subset of state of the art metrics with a very simple yet provably correct streaming algorithm.

To know more:

• [TNSM-21a] Huet, Alexis and Saverimoutou, Antoine and Houidi, Zied Ben and Shi, Hao and Cai, Shengming and Xu, Jinchun and Mathieu, Bertrand and Rossi, Dario, "Deployable models for approximating web QoE metrics from encrypted traffic" In IEEE Transactions on Network and Service Management, pp.839-854, mar. 2021, DOI 10.1109/TNSM.2020.3037019 Journal Abstract Bibtex

  @article{DR:TNSM-21a,
    title = {Deployable models for approximating web QoE metrics from encrypted traffic},
    author = {Huet, Alexis and Saverimoutou, Antoine and Houidi, Zied Ben and Shi, Hao and Cai, Shengming and Xu, Jinchun and Mathieu, Bertrand and Rossi, Dario},
    journal = {IEEE Transactions on Network and Service Management},
    year = {2021},
    month = mar,
    vol = {18},
    issue = {1},
    pages = {839-854},
    doi = {10.1109/TNSM.2020.3037019},
    howpublished = {https://nonsns.github.io/paper/rossi21tnsm-a.pdf},
    note = {project=huawei},
    topic = {qoe-web}
  }
  

  Being on endpoints, Content Providers can easily evaluate end users’ web browsing quality of experience (web QoE) by accessing in-browser computed application-level metrics. Because of end-to-end traffic encryption, it is becoming considerably harder for Internet Service Providers (ISPs) to evaluate the web QoE of their customers, which is important for management purposes. In this paper, we propose data-driven machine learning techniques and exact flow-level algorithmic methods to infer well-known application-level web performance metrics (such as SpeedIndex and Page Load Time) from raw encrypted streams of network traffic. We prove the efficiency of our approach taking as input a unique dataset of more than 200,000 experiments, targeting a large set of popular pages (Alexa top-500), from probes from several ISPs networks, with different browsers (Chrome, Firefox) and viewport combinations. Results show that our data-driven models are not only accurate for several web performance metrics, but also feature the ability to generalize to previously unseen conditions. Furthermore, we discuss how our extremely lightweight flow-level method has a provable accuracy on a specific metric, and is thus of particular appeal from a deployment viewpoint