Adding a New Dimension to HTTP Adaptive Streaming Through Multiple-Source Capabilities

Abstract

Adaptive streaming protocols over HTTP (HAS), such as MPEG-DASH, have become the de-facto solutions to deliver video over the Internet. By avoiding buffer stalling, HAS increases end-user's Quality of Experience (QoE). These solutions always bind a client to one server. We propose to extend HAS capabilities to multiple-source streaming, which offers the opportunity to obtain higher QoE by exploiting expanded bandwidth and link diversity in distributed streaming infrastructures. We expose a pragmatic evolving DASH-compliant solution for dynamic adaptive Multiple-Source Streaming over HTTP (MS-Stream) that simultaneously utilizes several servers. We validated our approach using network profiles from DASH Industry Forum and Internet users. Our solution is compared to optimal -yet practically unfeasible-DASH solutions in multiple-source environments and over several QoE criteria. Results show the QoE gains of using MS-Stream against DASH; an online demonstration of our work is available.

PrivaTube

“…However, for edgeassisted video delivery from peers with less stable or reliable links, fetching from multiple sources with some level of redundancy is a strong asset. The PrivaTube streaming protocol extends a multiple-source adaptive streaming protocol, MS-Stream [11], fully compatible with DASH. MS-Stream is illustrated by Figure 2, while Figure 1 presents the complete workflow of our extension.…”

“…PrivaTube extends MS-Stream [11], a protocol for video streaming using multiple sources, compatible with the leading MPEG-DASH standard [53]. It ensures that the load on the core servers is minimized, and that the impact of faults is masked through redundancy.…”

Enabling Multi-Source Coded Downloads

“…Bruneau et al have proposed MS-streaming, a multiplesource streaming solution that splits video into multiple independent sub-streams and offers methods for bit-rate adaptation and server-switching [6]. Compared to optimal Dynamic Adaptive Streaming over HTTP (DASH) systems, MS-streaming can achieve up to a 74% mean bit-rate gain.…”

QOE enhancement through cost-effective adaptation decision process for multiple-server streaming over HTTP

“…Based on the work in [3], we propose a dynamic adaptive Multiple-Source Streaming over HTTP protocol (MS-Stream) as an evolution of HAS solutions (and more specifically DASH) that simultaneously uses several servers for the download of each video segment. MS-Stream clients request several servers to create and deliver independent sub-segments (also referred as sub-streams).…”