Improving User Experience when HTTP Adaptive Streaming Clients Compete for Bandwidth

Abstract

The last decade has witnessed a tremendous increase of high-definition video consumption over the Internet and the emergence of the HTTP adaptive streaming technique, designed to cope with highly varying delivery conditions. Despite all its inherent advantages, HAS suffers from weaknesses when competing for bandwidth. In this paper, we focus on improving the HAS behavior for several cases of bandwidth competition. We propose in the first part a dynamic traffic shaping method operating in two phases (transient and steady) to improve convergence speed when HAS competes with a greedy TCP flow. However, when competition takes place between several HAS clients, this solution sub-optimally uses the bandwidth. Therefore, the second part investigates HAS client collaboration, defining information to be exchanged between clients and a common set of rules to determine the best representation to be requested.