Video Streaming QoE Management via Encrypted Traffic Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network operators face challenges in accurately managing and measuring video streaming quality of experience due to limited access to video information, especially with end-to-end encryption, and the complexity of factors influencing user experience across various devices and content providers.
Innovation Solution
A system and method for managing video streaming Quality of Experience (QoE) that collects data on video buffer health, resolution, throughput, and device type, creates a model to analyze these factors, and determines a QoE score, which can be aggregated across network dimensions to notify providers of subpar experiences, using machine learning and heuristic processes to provide a holistic view of user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If end-to-end encryption is implemented for video streaming traffic, then security and privacy are improved, but network operators lose access to video information such as codecs, frame-rate, and position
Solution Approach 1:
The system performs preliminary actions by collecting and storing video metadata (codecs, frame-rate, resolution, duration) before encryption occurs, and by proactively monitoring network conditions and device characteristics. This allows the QoE model to be built with complete information beforehand, eliminating the need to decrypt traffic for analysis.
Solution Approach 2:
The patent introduces an intermediary QoE estimation system that operates at the network layer without requiring decryption. This intermediary system uses available metadata and network measurements to infer video quality metrics, acting as a mediator between encrypted traffic and quality assessment requirements.
2Measurement precision
If comprehensive video quality measurement is implemented across multiple devices and content providers, then QoE accuracy is improved, but system complexity increases
Solution Approach 1:
The patent creates a universal QoE estimation model that works across multiple device types (smartphones, tablets, PCs, Smart TVs) and content providers (Netflix, YouTube, Amazon Prime) using a single standardized framework. The model adapts to different devices and providers through configurable parameters rather than requiring separate complex systems for each.
Solution Approach 2:
The system handles device and provider variability by changing parameters within a unified model rather than creating separate models. Device characteristics, content provider attributes, and network conditions are represented as adjustable parameters that modify the core QoE estimation algorithm, simplifying the overall system architecture.
3Reliability
If real-time video quality monitoring is implemented, then user experience management is improved, but data processing requirements and computational load increase
Solution Approach 1:
The system implements partial monitoring by selecting only the most relevant QoE factors for real-time calculation based on current network conditions and device characteristics. Not all video parameters are continuously monitored at full resolution - the system adjusts the level of monitoring based on what is necessary for accurate QoE estimation in each context.
Solution Approach 2:
Computationally intensive tasks such as model training, parameter calibration, and baseline establishment are performed in advance during low-load periods. This preliminary processing reduces the computational burden during real-time QoE monitoring, allowing the system to make rapid assessments without excessive processing demands.
Data Source
AI summary
A method for managing Quality of Experience (QoE) for video streaming traffic flow on a network, the method including: collecting data associated with a plurality of video streaming traffic flows; creating a model based on the collected data; determining factors associated with a new video streaming traffic flow; analyzing the factors based on the model; determining a QoE score based on the analysis. A system for managing QoE for video streaming traffic flow on a network, the system including: a factor determination module configured to collect data associated with a plurality of video streaming traffic flows; a model module configured to create a model based on the collected data; an analysis module configured to determine factors associated with a new video streaming traffic flow and analyze the factors based on the model; and a QoE engine configured to determine a QoE score based on the analysis.