Multimedia Networking
optimizing quality and user experience of video streaming services
Video streaming services dominate Internet traffic. It is therefore essential to enhance user experience, particularly for new types of streaming services, over new wireless access systems, and through emerging learning-based methods.
Selected papers
[JSAC] Enabling Robust DRL-driven Networking Systems via Teacher-Student Learning
Ying Zheng, Lixiang Lin, Tianqi Zhang, Haoyu Chen, Qingyang Duan, Yuedong Xu*, Xin Wang
IEEE JSAC 2021.
We introduced a teacher–student paradigm to enhance the robustness of deep reinforcement learning (DRL) in networking systems, where the teacher is an imperfect but domain-specific algorithm, and the student is a generally well-performing DRL agent.
[ToN/Infocom] Enabling Practical Large-Scale MIMO in WLANs With Hybrid Beamforming
Zhe Chen, Xu Zhang, Sulei Wang, Yuedong Xu*, Jie Xiong, Xin Wang
IEEE/ACM Trans. Networking, 2021.
[TMM] MuVi: Multi-view Video Aware Transmission over MIMO Wireless Systems
Zhe Chen, Xu Zhang, Yuedong Xu*, Xin Wang, Jie Xiong and Yu Zhu
IEEE Trans. Multimedia, 2017.
We proposed novel resource allocation schemes for MIMO WiFi systems and 3D video streams. In particular, we integrated MIMO angular localization with hybrid beamforming to effectively mitigate the excessive overhead of CSI feedback. In the first round, wireless clients transmit preambles to the access point, enabling it to estimate the angles of arrival and perform proper client selection. In the second round, the clients provide fine-grained CSI feedback, allowing the access point to carry out conventional precoding and beamforming operations. These studies further inspired our subsequent work on WiFi localization.
[TMC] Modeling Buffer Starvations of Video Streaming in Cellular Networks with Large-Scale Measurement of User Behavior
Yuedong Xu, Zhujun Xiao, Hui Feng, Tao Yang, Bo Hu and Yipeng Zhou
IEEE Trans. Mobile Computing, 2017.
[TMC/Infocom] Flow Level QoE for Video Streaming in Wireless Networks
Yuedong Xu, Salah Elayoubi, Eitan Altman, Rachid El-Azouzi
IEEE Trans. Mobile Computing, 2016.
[TMC] Analytical QoE Models for Bit-rate Switching in Dynamic Adaptive Streaming Service
Yuedong Xu, Yipeng Zhou, Dah-ming Chiu
IEEE Trans. Mobile Computing, 2014.
[TMM/Infocom] Analysis of Buffer Starvation with Application to QoE Optimization of Streaming-like Services
Yuedong Xu, Eitan Altman, Rachid El-Azouzi, Majed Haddad, Salah Elayoubi, Tania Jimenez
IEEE Trans. Multimedia, 2014.
Predicting the probability of playback interruption is essential for enhancing the quality of user experience. We develop an analytical framework that reformulates the playback interruption problem as a risk analysis task in actuarial science. Interestingly, actuarial science and video QoE are two distinct domains that nevertheless share several common analytical tools.