Optimizing Uplink Bandwidth Utilization for Crowdsourced Livecast

Driven by the prevalence of video generation devices and the development of network infrastructures, there has been an explosive growth of Crowdsourced Video Livecast (CVL) services in the past few years. Significant efforts have been made to provide high quality CVL services with limited bandwidth availability. However, most of the existing works focused on optimizing downlink bandwidth for video distribution rather than uplink bandwidth for video uploading. For example, uploaders (i.e., broadcasters) in Twitch can arbitrarily set their upload rates, which may lead to a significant waste of upload bandwidth with the increasing number of uploaders. In this paper, we propose an effective low-complexity algorithm called Baal to optimize upload bandwidth allocation among massive uploaders. Our objective is to optimize the utility of video uploading from the perspective of CVL platform operators by considering both viewers Quality-of-Experience (QoE) and upload bandwidth cost. To guarantee the effectiveness and fairness of bandwidth allocation, we adopt the optimization framework of Nash Bargaining Solution (NBS), which can determine the optimal bandwidth budget, upload bitrate and datacenter selection for each uploader jointly. Finally, we conduct extensive trace-driven simulations to evaluate our proposed algorithm and the results show that our algorithm achieves much higher utility than alternative strategies in various conditions.