Fingerprint-Based Web Object Identification for Caching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Over-the-top (OTT) media traffic consumes significant bandwidth in ISP networks, leading to increased transit costs and impacting customer experience, as traditional caching methods fail due to non-cacheable HTTP properties and lack of explicit relationships between ISPs and OTT content providers.
Innovation Solution
Implementing a transparent proxy caching (TPC) engine that calculates a unique fingerprint for content objects based on their payload and metadata, storing it in a content manager database, and using a mesh of TPC engines to accelerate content delivery by caching and relaying fingerprints between engines without requiring pairing relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional caching methods are used for OTT content, then cache management is simple, but cache-hit ratio is low because OTT content has non-cacheable HTTP properties and cannot be uniquely identified through URLs
Solution Approach 1:
The patent introduces a fingerprint as an intermediary identifier between the OTT content and the caching system. Instead of relying on traditional URL-based identification which fails for OTT content, the system generates a unique fingerprint from the content payload itself. This fingerprint serves as a mediator that enables reliable content identification and caching without requiring changes to the OTT content delivery protocol or relying on URL properties that don't exist for streamed content.
Solution Approach 2:
The patent replaces the traditional HTTP-based identification mechanism (which relies on URLs, cache-control headers, and other HTTP properties) with a fingerprint-based identification system. The fingerprint is generated by processing the actual content payload through a hash function, substituting the mechanical HTTP identification system with a content-based identification approach that works independently of HTTP caching semantics.
2Loss of energy
If OTT content is cached without unique identification, then bandwidth consumption is high, but implementing traditional caching increases transit costs due to lack of explicit relationships between ISPs and OTT content providers
Solution Approach 1:
The patent enables the OTT content itself to provide its own identification mechanism through the fingerprint. The content payload is processed to generate a unique identifier that the caching system can use autonomously. This self-service approach eliminates the need for explicit relationships or coordination between ISPs and OTT content providers, as the fingerprinting mechanism works independently of any external agreements or protocol modifications.
3Productivity
If a mesh of TPC engines is implemented to accelerate content delivery, then delivery efficiency improves, but system complexity increases due to multiple engines and fingerprint relay mechanisms
Solution Approach 1:
The patent divides the content delivery function into multiple independent TPC engines distributed across the network. Each engine operates autonomously, generating and caching fingerprints locally without requiring centralized coordination. This segmentation allows parallel processing of content requests and enables the mesh topology to accelerate delivery through multiple paths while maintaining independent operation of each node.
Solution Approach 2:
The patent implements preliminary fingerprinting where TPC engines pre-generate and cache fingerprints for OTT content before actual content delivery is needed. When content is requested, the fingerprint is already available in the cache, enabling immediate identification and delivery without real-time processing delays. This preliminary action eliminates the need for complex real-time coordination between mesh engines during content delivery.
Data Source
AI summary
In some embodiments, a request for a content object may be received as a data packet. Upon determining that the content object is not cached, an element of the data packet may be modified and the modified data packet may be forwarded toward a content provider of the content object.