Dynamic CDN Streamer Selection via Real-Time Popularity

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Content delivery networks (CDNs) face inefficiencies due to static streamer selection processes that do not account for real-time content popularity, leading to reduced Cached Hit Ratio, increased response times, high traffic to Origin Servers, and poor user experience, especially for Video on Demand content with low popularity and concurrency.

Innovation Solution

A dynamic method for selecting CDN streamers and resources based on real-time content popularity, where streamers calculate audience data and send it to a Request Router for aggregation, allowing for differentiated thresholds and traffic concentration policies to optimize streamer selection and reduce network traffic.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Area of stationary object

If content is distributed over all available PoPs according to network topology, then network coverage is improved, but Cached Hit Ratio decreases due to low popularity content being distributed widely

Engineering Contradiction:
Improvenetwork coverageVSAvoidCached Hit Ratio
Core Design Contradiction:
Area of stationary objectVSReliability

Solution Approach 1:

The patent applies local quality by differentiating content distribution strategies based on popularity metrics. Popular content is distributed across multiple PoPs to maximize cache hit ratios, while non-popular content is concentrated in fewer PoPs. This localized adaptation of distribution strategy to content characteristics resolves the contradiction between wide coverage and high cache efficiency.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system dynamically adjusts content distribution based on real-time popularity metrics and demand patterns. The request router continuously monitors content performance and repositions content between PoPs accordingly, transforming the static distribution model into a dynamic one that adapts to changing popularity, thereby maintaining both coverage and cache hit ratio.

Inventive Principle:
Principle #15Dynamics

2Area of stationary object

If non-popular content is distributed across many PoPs, then network coverage is improved, but response time increases due to cache misses

Engineering Contradiction:
Improvenetwork coverageVSAvoidresponse time
Core Design Contradiction:
Area of stationary objectVSLoss of time

Solution Approach 1:

The patent implements local quality by applying different distribution strategies to different content segments. Non-popular content is concentrated in specific PoPs with appropriate caching resources, while popular content is widely distributed. This localized optimization reduces cache misses for non-popular content without sacrificing overall network coverage.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system performs preliminary actions by pre-positioning non-popular content in selected PoPs based on predicted demand patterns and popularity metrics. This advance placement ensures that when requests for non-popular content arrive, the content is already available in optimal locations, reducing response time while maintaining coverage.

Inventive Principle:
Principle #10Preliminary action

3Area of stationary object

If content is widely distributed across PoPs, then network coverage is improved, but traffic to Origin Servers increases due to cache misses

Engineering Contradiction:
Improvenetwork coverageVSAvoidtraffic to Origin Servers
Core Design Contradiction:
Area of stationary objectVSLoss of energy

Solution Approach 1:

The patent applies local quality by optimizing content placement at edge PoPs based on popularity and demand characteristics. Non-popular content is strategically positioned in PoPs with available caching capacity, reducing the need to fetch content from origin servers. This localized caching strategy reduces origin server traffic while preserving network coverage.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system creates copies of non-popular content in selected PoPs based on demand predictions and popularity metrics. These local copies serve requests without requiring origin server communication, reducing traffic to origin servers while maintaining the ability to deliver content across the network.

Inventive Principle:
Principle #26Copying

4Device complexity

If streamer selection is based on static criteria, then device complexity is reduced, but CDN efficiency decreases due to inability to adapt to content popularity

Engineering Contradiction:
Improvestreamer selection processVSAvoidCDN efficiency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent implements feedback mechanisms where the request router continuously monitors content popularity metrics, cache hit ratios, and demand patterns. This feedback information is used to dynamically adjust streamer selection and content distribution decisions, enabling the system to adapt to changing popularity without requiring complex manual configuration or static rules.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs self-service by automatically monitoring its own performance metrics and making autonomous decisions about content distribution and streamer selection. The request router uses real-time popularity data to dynamically optimize content placement and delivery decisions without external intervention, improving CDN efficiency while keeping the control logic relatively simple.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3879798B1Method for dynamic selecting delivery resources and policies in content delivery networks based on content popularity
Publication Date: 2025.03.26 TELEFONICA SA
  • EP3879798B1 patent drawingFigure 1
  • EP3879798B1 patent drawingFigure 2
  • EP3879798B1 patent drawingFigure 3(A)~3(B)

AI summary

A method for selecting delivery resources/policies in a CDN wherein an optimal PoP is assigned to an identified content requested by a user, comprising: - for each identified content and the assigned optimal PoP, calculating audience information in real time; - recollecting, e.g., by a Request Router (100), the calculated audience information from all the PoPs or sites (Sx, Sy) serving the content to obtain global audience information; - defining at least an audience threshold per content and PoP; - determining a degree of popularity of the content by comparing the calculated audience information or the obtained global audience information with the audience threshold; and - selecting a delivery resource, e.g., a streamer (300x, 300y), and/or a delivery policy according to a popularity policy comprising rules based at least on the degree of popularity for the requested content.