A comprehensive analytical framework for VoD services in hybrid CDN-P2P systems. (1st July 2020)

• Record Type: Journal Article
• Authors: Torres-Cruz, Noé; Rivero-Angeles, Mario E.; Rubino, Gerardo; Menchaca-Mendez, Ricardo; Menchaca-Mendez, Rolando; Ramirez, David
• Is Part Of: Journal of network and computer applications. Volume 161(2020)
• Publication Date: 2020-07-01
• Subjects: Video streaming -- Video on demand -- Hybrid CDN-P2P system -- Fluid model -- Markov-chain model -- Resource allocation schemes -- GPWA; Microcomputers -- Periodicals; Computer networks -- Periodicals; Application software -- Periodicals; Micro-ordinateurs -- Périodiques; Réseaux d'ordinateurs -- Périodiques; Logiciels d'application -- Périodiques; Application software; Computer networks; Microcomputers; Periodicals; 004.05; 004
• DOI: 10.1016/j.jnca.2020.102643
• Languages: English
• ISSNs: 1084-8045

Abstract

Abstract: Video on Demand services generate the largest amount of traffic in networks nowadays. Previous works have proposed integrating Content Delivery Networks (CDN) and Peer-to-Peer (P2P) networks to satisfy this demand. However, their analytical methods do not consider all the factors that affect the performance of these systems. Hence, we present a novel comprehensive framework, based on a fluid model, to evaluate VoD services over hybrid CDN-P2P systems. The proposed framework considers features of the service (e.g., size, coding-rate and popularity of the video), network attributes (e.g., upload data-rate capacity of servers and peers) and characteristics of the behavior of the users (e.g., sojourn time, cooperativeness and frequency of random-seeks). Our framework allows a system to be evaluated under a wide variety of scenarios in terms of network-cost and Quality-of-Service (QoS) parameters and is flexible enough to model different resource allocation schemes, including a novel scheme to distribute the upload network capacity. Despite the wide variety of considered factors, our framework is tractable and as accurate as discrete-models based on Markov chains. Highlights: VoD services over hybrid P2P-CDN systems are evaluated through fluid models and Markov chains. The effects of arrivals, departures, video-file sharing, playing and VCR functions are analyzed. The models permit evaluating multiple resource allocation schemes, including a novel proposal. The system is evaluated in terms of network-costs (P2P network and CDN), and QoS parameters. The developed framework is flexible enough to evaluate the system in a wide variety of scenarios.