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

TL;DR

Countermeasures against piracy are introduced, 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.

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.

Figures (4)

• Figure 1: Number of NVD tags in GitHub hyperlinks, over time (semesters)
• Figure 2: Proposed framework for detecting patches and exploit on GitHub pages.
• Figure 3: (a) Initial dataset comprising 38,180 events (33,702 unique URLs) cited by NVD; (b) NVD patch and exploit tags after each rule propagation step.
• Figure 4: Decision tree trained in the task of predicting if a link contains a patch.