Packet Flow Categorization via Timing Pattern Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current communication networks struggle to differentiate and provide high-quality delivery for adaptive bitrate (ABR) content flows, especially when encrypted with Digital Rights Management (DRM), as existing techniques like Deep Packet Inspection (DPI) are ineffective in such scenarios, leading to unsatisfactory service experiences for consumers and increased burdens on network operators.
Innovation Solution
A system and method for categorizing packet flows in a network node by monitoring packets and determining specific traffic patterns associated with ABR flows, allowing for policy-based operations, even when content is encrypted, using a combination of packet tracking and pattern recognition modules within a Software-Defined Network (SDN) framework.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Deep Packet Inspection (DPI) is used to identify ABR flows, then flow recognition accuracy is improved, but computational complexity and processing overhead increase significantly
Solution Approach 1:
The patent extracts only the essential timing characteristics (inter-arrival times and packet sizes) from packets, rather than performing full DPI inspection of packet contents. This extraction approach maintains flow recognition accuracy while dramatically reducing computational complexity by focusing only on timing patterns visible in packet headers.
Solution Approach 2:
The patent segments the flow identification process into discrete timing measurements (inter-arrival times) and packet size observations, comparing these segmented characteristics against reference patterns. This segmentation enables efficient processing by breaking down complex flow analysis into simple, repeatable timing comparisons.
2Reliability
If ABR content is encrypted using DRM techniques, then content security is improved, but network node ability to identify and differentiate ABR flows deteriorates
Solution Approach 1:
The patent performs preliminary actions by establishing reference ABR flow patterns before encrypted traffic arrives. These reference patterns capture the timing and size characteristics of ABR flows, enabling network nodes to identify encrypted ABR traffic by comparing incoming packet timing against pre-established references, thus maintaining identification capability despite encryption.
3Reliability
If traditional linear TV delivery model is used, then service quality control is improved, but adaptability to modern consumption patterns deteriorates
Solution Approach 1:
The patent introduces dynamics by enabling network nodes to adaptively recognize and categorize different types of traffic flows (ABR, linear TV, web browsing) based on observed packet timing patterns. This dynamic flow categorization allows the network to apply appropriate quality control policies to different traffic types, maintaining service quality control while adapting to diverse consumption patterns including on-demand ABR streaming.
Data Source
AI summary
A scheme for of categorizing packet flows through a network node, e.g., a proxy or router. In one embodiment, packets of a particular flow arriving at the node are monitored for determining whether there is a specific packet traffic pattern associated with the particular flow. Responsive to the determining, appropriate techniques may be utilized for recognizing the specific packet traffic pattern as belonging to a category of packet flow, e.g., ABR video flow.