A Dynamic Live Streaming Service Architecture Integrated Sensing and Control

Abstract

AbstractThe live streaming of Over-The-Top (OTT) hot events will cause the tidal traffic of the network. As viewership grows, this high network load can lead to network blocking, poor quality of experience for users, high transmission costs for operators, and increased energy consumption. Most OTT lives content providers offer video services with unicast. To save bandwidth, there are some transmission solutions that also use multicast. Transmitting OTT video using the existing Internet Protocol television (IPTV) network is a win-win situation for telecom operators and OTT content providers. However, so far as we know, there is no dynamic service system for live streaming content that is easy to control over the existing IPTV network. As a result, it is more difficult to provide OTT live content on the existing IPTV network. In this paper, we propose a dynamic system architecture for live streaming services with integrated sensing and control. This architecture can leverage existing IPTV networks for real-time smart sensing and control of media services. The client does not need to make modifications and still uses the unicast protocol for video content delivery. We use a MA as a bridge between multicast and unicast. It receives client requests, converts multicast to unicast, and provides services to users. In addition, we design a unicast-multicast convergence method, which first provides unicast services to viewers and then dynamically switches to multicast service, reducing bandwidth and also improving the viewer experience.KeywordsMulticastUnicastOTT VideoLive streaming