Abstract

The rapid development of content delivery networks and cloud computing has facilitated crowdsourced live-streaming platforms (CLSP) that enable people to broadcast live videos which can be watched online by a growing number of viewers. However, in order to ensure reliable viewer experience, it is important that the viewers should be provided with multiple standard video versions. To achieve this, we propose a joint fog-assisted transcoding and viewer association technique which can outsource the transcoding load to a fog device pool and determine the fog device with which each viewer will be associated, to watch desired videos. The resulting non-convex integer programming has been solved using a computationally attractive complementary geometric programming (CGP). The performance of the proposed algorithm closely matches that of the globally optimum solution obtained by an exhaustive search. Furthermore, the trace-driven simulations demonstrate that our proposed algorithm is able to provide adaptive bit rate (ABR) services.

Highlights

  • The current crowdsourced live streaming has allowed a growing number of people start to broadcast live videos via the crowdsourced live-streaming platforms (CLSP) such as Twitch.TV, Douyu.TV, and Periscope

  • When an original live video is uploaded by a broadcaster i to the regional data centers, the CLSP scheduler will decide which representation should be transcoded by which fog device based on the viewer requirements

  • All the standard transcoded live videos will be transmitted from the fog devices to the associated viewers

Read more

Summary

INTRODUCTION

The current crowdsourced live streaming has allowed a growing number of people start to broadcast live videos via the CLSPs such as Twitch.TV, Douyu.TV, and Periscope. In Twitch.TV, only the premium broadcasters have access to the ABR service, and for the rest of the broadcasters, only the original versions are available The reason behind this is that a general cloud instance can only deal with at most two transcoding tasks simultaneously. Most of CLSPs support the broadcasters and viewers with interactive chat service Under such a scenario, the latency problem has become even more significant than the traditional live streaming platforms. In [10], a case study is presented for Twitch.TV This case demonstrates that with the advance of personal computing devices, a significant fraction of CLSP viewers potentially has appropriate computing resources for stable real-time transcoding. To the best of our knowledge, this paper is the first work to consider a joint transcoding task assignment and viewer ƌŽĂĚĐĂƐƚĞƌϭ ƌŽĂĚĐĂƐƚĞƌϮ ƌŽĂĚĐĂƐƚĞƌϯ ƌŽǁĚƐŽƵƌĐĞĚůŝǀĞƐƚƌĞĂŵ ƉůĂƚĨŽƌŵ;>^WͿ ^ĐŚĞĚƵůĞ ŝŶĨŽƌŵĂƚŝŽŶ

SYSTEM MODEL
Cost model
QoE model
Network utility
Problem formulation
Proposed algorithm
Overhead analysis
Exhaustive search
Trace-driven simulation
Broadcaster3
CONCLUSION
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call