Cloud-Based Internet Linear Video Delivery System Using Edge Computing on a CDN

Daiki Fukudome, Satoshi Nishimura

Internet users are increasingly exposed to linear streaming services. In several linear delivery platforms that use Hypertext Transfer Protocol (HTTP) adaptive streaming (HAS), programs are scheduled in a manipulated manifest file as a playlist. However, when an urgent program is dynamically inserted, like television broadcasting, ensuring reliable and rapid program switching is difficult, especially in low-latency delivery, because all the players' manifest files must be quickly updated. Thus, we propose a cloud-based HAS linear delivery system that facilitates prompt schedule changes by utilizing edge computing within the content delivery network (CDN). We propose an end-to-end system that focuses on the efficient use of cloud resources for encoding and program switching at pre-scheduled times by leveraging the CDN's edge capabilities and simple low-latency playback. Moreover, the prompt program schedule changes are addressed, particularly in scenarios involving dynamic program insertions and the replacement of certain programs based on viewers' attributes on CDN edge.

Print ISSN
Electronic ISSN
2160-2492
Published
2024-10
Content type
Original Research
Keywords
linear streaming, http adaptive streaming (has), dynamic adaptive streaming over http (dash)
DOI
10.5594/JMI.2024/TPSA1162