Encrypted Video Traffic Segmentation for Accurate QoE Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge of analyzing video quality of experience is exacerbated by the high share of encrypted video traffic, which introduces biases and inaccuracies in key performance indicators due to the inability to effectively segment audio and video streams, particularly under poor network conditions.
Innovation Solution
A method and system for segmenting audio and video requests in encrypted video streams using partial request-response reconstruction and adaptive linear discriminant functions to separate audio and video traffic, enabling accurate quality of experience (QoE) metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If video traffic analysis is performed on encrypted streams, then privacy protection is maintained, but measurement precision of QoE metrics deteriorates
Solution Approach 1:
The patent segments encrypted video traffic into distinct audio and video components using machine learning classification. By analyzing traffic patterns, packet sizes, and timing characteristics without decrypting the content, the system separates audio streams from video streams, enabling accurate QoE measurement while preserving encryption and privacy.
Solution Approach 2:
The patent introduces an intermediary classification layer between the encrypted traffic and QoE analysis. This intermediary uses machine learning models to interpret encrypted traffic characteristics and extract meaningful QoE metrics without requiring access to the actual encrypted content, thus maintaining privacy while enabling measurement.
2Device complexity
If audio and video streams are not segmented, then analysis complexity is reduced, but measurement precision of QoE metrics deteriorates due to biases
Solution Approach 1:
The patent automatically segments mixed audio-video traffic into separate streams using machine learning classification. This segmentation enables independent QoE analysis of each stream, eliminating measurement biases that would occur if audio and video were analyzed together as a single mixed stream.
3Ease of operation
If traditional traffic classification is used, then ease of operation is maintained, but measurement precision deteriorates under poor network conditions
Solution Approach 1:
The patent employs dynamic machine learning classification that adapts to varying network conditions. The system continuously learns from traffic patterns and adjusts its classification parameters in real-time, maintaining high measurement precision even when network conditions change or deteriorate, unlike static traditional classification methods.
Solution Approach 2:
The patent changes the parameters used for traffic classification from simple header-based rules to complex multi-dimensional features including packet size distributions, timing patterns, and flow characteristics. These parameter changes enable more accurate discrimination between audio and video streams under diverse network conditions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
We generally describe a method (700) for analyzing data traffic quality of data traffic flowing between a user equipment and a server of a video service provider. The data traffic comprises video traffic and audio traffic. The method comprises separating (S702), in the data traffic, the video traffic from the audio traffic, and analyzing (S704), based on the separated video traffic, the data traffic quality of the data traffic flowing between the user equipment and the server of the video service provider.