Unsupervised methodology to unveil content delivery network structures
US9686173B1 · kind B1 · utility
Assignee
Inventors
Key dates
| Filing date | Oct 27, 2014 |
| Grant date | Jun 20, 2017 |
| Priority date | — |
| Expiry date | Jun 26, 2035 |
Classification
- Technology area (CPC H)Electricity
- CPC primaryH04L43/0876
- WIPO fieldDigital communication
- WIPO sectorElectrical engineering
Abstract
A method for analyzing a content delivery network. The method includes obtaining network traffic flows corresponding to user nodes accessing contents from a set of servers of the content delivery network, extracting a timing attribute from each network traffic flow associated with a server, where the timing attribute is aggregated into a timing attribute dataset of the server based on all network traffic flows associated with the server, generating a statistical measure of the timing attribute dataset as a portion of a feature vector representing the server, where the feature vector is aggregated into a set of feature vectors representing the set of servers, analyzing the set of feature vectors based on a clustering algorithm to generate a set of clusters, and generating, based on the set of clusters, a representation of server groups in the content delivery network.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.