Network Node QoE Estimation for Video Streaming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Internet Service Providers (ISPs) face challenges in identifying and addressing issues that negatively impact customers' Quality of Experience (QoE) related to streaming media consumption on personal electronic devices without requiring access to the devices or generating additional network traffic.
Innovation Solution
A passive data collection method that uses network nodes to monitor and estimate objective QoE metrics, such as streaming resolution changes, buffering, and start-up delays, allowing ISPs to proactively diagnose and adjust network parameters to improve QoE without customer feedback.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ISPs use traditional QoE monitoring methods requiring client device access, then measurement accuracy is improved, but device complexity and user privacy intrusion increase
Solution Approach 1:
The patent introduces network nodes as intermediaries between content servers and client devices. These nodes passively monitor network traffic parameters (packet sizes, inter-arrival times, retransmission rates) to infer QoE metrics without requiring direct access to or modification of client devices. The network node acts as a mediator that collects data from the network layer rather than requiring device-level instrumentation.
Solution Approach 2:
The patent replaces the mechanical approach of installing software agents or hardware components on client devices with a network-based statistical inference system. Instead of directly measuring QoE parameters at the device level, the system uses network traffic characteristics as proxies and applies machine learning models to predict QoE metrics, eliminating the need for client device modifications.
2Difficulty of detecting and measuring
If ISPs implement active QoE monitoring systems, then QoE detection capability is improved, but network traffic increase and system complexity worsen
Solution Approach 1:
The system uses the network's existing traffic data for QoE monitoring purposes, making the network self-service for quality assessment. The network nodes utilize their normal packet inspection functions to collect data for QoE estimation without requiring separate monitoring infrastructure or additional traffic generation. The existing network infrastructure serves dual purposes: data transmission and quality measurement.
Solution Approach 2:
The patent replaces active monitoring mechanisms that would generate additional network traffic with passive statistical analysis of existing traffic patterns. Instead of actively probing or injecting monitoring traffic, the system analyzes natural network traffic characteristics (packet sizes, timing, retransmissions) to infer QoE metrics, eliminating the need for separate monitoring traffic streams.
3Measurement precision
If ISPs collect detailed QoE data from network traffic, then QoE estimation accuracy is improved, but data processing complexity and computational resources increase
Solution Approach 1:
The patent segments the QoE estimation process into distinct functional components: data collection at network nodes, feature extraction from traffic patterns, and QoE metric prediction using machine learning models. This segmentation allows each component to be optimized independently and enables distributed processing where network nodes handle data collection and preliminary feature extraction, while centralized systems perform complex modeling operations.
Solution Approach 2:
The system transforms raw network traffic parameters (packet sizes, inter-arrival times, retransmission counts) into derived features that are more predictive of QoE metrics. By changing the representation of data from raw traffic parameters to processed features through statistical analysis and machine learning, the system improves estimation accuracy while managing computational complexity through efficient feature transformation rather than complex processing at every stage.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Computer-implemented systems and methods for diagnosing and correcting connectivity problems in a network are provided. The method includes obtaining, at a network node disposed at an edge of the network, measurements of one or more attributes of network traffic exchanged between a content server and a client device via the network. The network traffic is associated with streaming video content transmitted from the content server to the client device over the network. The method also includes analyzing the one or more attributes of the network traffic to estimate Quality of Experience (QoE) performance metrics related to the streaming video content at the client device, and adjusting one or more network operating parameters of the network responsive to QoE performance metrics falling below a predetermined threshold.