An Optimal Prefix Replication Strategy for VoD Services

TL;DR

This paper introduces a scalable proxy server architecture and an optimal video prefix replication strategy based on regional popularity and scene changes, aiming to improve video availability, reduce latency, and enhance request servicing in VoD services.

Contribution

It presents a novel regional popularity-based video prefix replication strategy combined with a scene change caching algorithm, optimizing video delivery in VoD systems.

Findings

Increased video availability and request servicing rate

Reduced client rejection ratio and startup latency

Effective utilization of Zipf-like popularity distribution

Abstract

In this paper we propose scalable proxy servers cluster architecture of interconnected proxy servers for high quality and high availability services. We also propose an optimal regional popularity based video prefix replication strategy and a scene change based replica caching algorithm that utilizes the zipf-like video popularity distribution to maximize the availability of videos closer to the client and request-servicing rate thereby reducing the client rejection ratio and the response time for the client. The simulation results of our proposed architecture and algorithm show the greater achievement in maximizing the availability of videos, client request-servicing rate and in reduction of initial start-up latency and client rejection ratio.