Minimally Intrusive Access Management to Content Delivery Networks based on Performance Models and Access Patterns

TL;DR

This paper proposes a minimally intrusive access management framework for CDNs that uses performance models and access patterns to detect misuse, combat piracy, and improve infrastructure efficiency.

Contribution

It introduces a novel approach combining performance analysis, statistical access patterns, and countermeasures to enhance CDN security and efficiency with minimal intrusion.

Findings

Defined acceptable request limits to detect abnormal access

Implemented quality degradation for pirate users as a countermeasure

Used queuing models to evaluate system performance under piracy scenarios

Abstract

This paper presents an approach to managing access to Content Delivery Networks (CDNs), focusing on combating the misuse of tokens through performance analysis and statistical access patterns. In particular, we explore the impact of token sharing on the content delivery infrastructure, proposing the definition of acceptable request limits to detect and block abnormal accesses. Additionally, we introduce countermeasures against piracy, such as degrading the quality of service for pirate users to discourage them from illegal sharing, and using queuing models to quantify system performance in different piracy scenarios. Adopting these measures can improve the consistency and efficiency of CDN access and cost management, protecting the infrastructure and the legitimate user experience.