Insights from Analysis of Video Streaming Data to Improve Resource Management
Sabidur Rahman,Hyunsu Muny,Hyongjin Leey,Youngseok Lee,Massimo Tornatore,Biswanath Mukherjee
DOI: https://doi.org/10.48550/arXiv.1806.08516
2018-09-06
Abstract:Today a large portion of Internet traffic is video. Over The Top (OTT) service providers offer video streaming services by creating a large distributed cloud network on top of a physical infrastructure owned by multiple entities. Our study explores insights from video streaming activity by analyzing data collected from Korea's largest OTT service provider. Our analysis of nationwide data shows interesting characteristics of video streaming such as correlation between user profile information (e.g., age, sex) and viewing habits, viewing habits of users (when do the users watch? using which devices?), viewing patterns (early leaving viewer vs. steady viewer), etc. Video on Demand (VoD) streaming involves costly (and often limited) compute, storage, and network resources. Findings from our study will be beneficial for OTTs, Content Delivery Networks (CDNs), Internet Service Providers (ISPs), and Carrier Network Operators, to improve their resource allocation and management techniques.
Networking and Internet Architecture