VAMP: A Predictive Approach to Audio/Video Bitrate Adaptation Over Wireless Networks

Adaptive bitrate (ABR) algorithms are employed for delivering media content across wireless networks. Current ABR schemes only focus on video bitrate adaptation, considering that audio content encoding has negligible impact on streaming quality, due to its smaller size. However, many commercial platforms use high-quality audio which is significant in size. Improper audio bitrate selection thus leads to imbalanced audio/video buffer levels and adversely affects video bitrate selection. To tackle this issue, we propose VAMP, an approach that drives Video and Audio adaptation using Multi-step Predictions. Experimental results demonstrate VAMP’s vast superiority over SOTA solutions.