Optimizing ABR and intelligent transport protocol selection

Abstract

The ongoing research delves into the intelligent transport protocol selection mechanism and bit-rate optimization on the face of protocol switching to enhance Quality of Experience (QoE) in dynamic network conditions. QoE is a metric for video streaming platforms, where the performance of ABR algorithms plays a pivotal role. The primary function of an ABR algorithm is to determine the quality of video segments requested from the server. Both overestimation and underestimation of video quality can significantly degrade QoE, necessitating more precise and adaptive decision-making mechanisms. The study aims to improve existing ABR algorithms, such as BOLA and Pensieve, by intro- ducing enhancements that account for dynamic network conditions. A key innovation involves leveraging both QUIC and TCP protocols in tandem, enabling DASH player to dynamically select the optimal protocol for segment requests based on network conditions. While TCP out- performs QUIC in high-bandwidth scenarios, video streaming platforms often continue to rely solely on QUIC, potentially missing opportunities for performance gains. Our experiments explore the impact of protocol switching on ABR performance in both low- and high-bandwidth environments. Preliminary tests demonstrate that intelligent protocol switching can significantly enhance ABR performance by optimizing segment requests.