An Efficient Approach for (Multi-)CDN Identification
M. B. Zhou, Jiaqi Zheng, Guihai Chen, Wanchun Dou · 2023
As the use of Content Delivery Networks (CDNs) continues to grow, it has become increasingly difficult to accurately identify which CDNs are being used by a particular website or domain. In this paper, we propose a new method for detecting CDNs, including multiple-CDN deployments, behind domains using a combination of DNS and HTTP-based measurements. Our key observation is that using a few vantage points can detect nearly all CDN(s) behind the websites. Thus, we first design a heuristic algorithm to select representative vantage points and retrieve DNS and HTTP(S) information to identify patterns that are indicative of the use of CDNs. We run our method on the top 10k websites in tranco list and detect 6,392 domains hosted on CDNs, of which 200 are hosted on multiple-CDN, accounting for 2.0% of the total.