Hybrid Fiber Coax Network Analyzer for Quality of Experience Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies fail to effectively predict and maintain customer quality of experience (QoE) in hybrid fiber coaxial networks, particularly due to factors like delay, jitter, and packet loss, which impact the satisfaction of subscribers with multimedia services such as IPTV, as they cannot constantly poll subscribers for feedback.
Innovation Solution
A method and system for processing quality of service data by varying network parameter states, obtaining subscriber feedback, statistically analyzing it to identify correlations, and producing scores representing predicted quality of experience, utilizing a headend with a network analyzer in a hybrid fiber/coax (HFC) network to provide a prediction of QoE.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If subscribers are constantly polled for feedback to accurately measure quality of experience, then measurement precision of QoE is improved, but loss of time and increased complexity of the system occur
Solution Approach 1:
The system performs preliminary actions by collecting and storing subscriber feedback at specific intervals rather than continuously polling. Feedback is collected during natural interaction moments (e.g., when services are actually used) and stored for later analysis, eliminating the need for constant real-time polling while maintaining measurement accuracy.
Solution Approach 2:
The system implements a feedback mechanism where subscriber responses to service quality questions are collected, stored in a database, and used to train machine learning models. These models then predict QoE metrics based on network parameters, creating a closed-loop system that improves measurement efficiency through iterative learning rather than continuous direct polling.
2Reliability
If network parameters are continuously monitored to predict quality of experience, then reliability of QoE prediction is improved, but device complexity and energy consumption increase
Solution Approach 1:
The system introduces machine learning models as intermediary components between raw network parameter data and QoE predictions. These models are trained offline using historical feedback data and then deployed to efficiently predict QoE based on current network parameters, reducing the complexity of real-time analysis while maintaining high prediction reliability.
Solution Approach 2:
The machine learning models are trained in advance using historical network parameter data and subscriber feedback. This preliminary training phase allows the models to learn complex relationships between network parameters and QoE, enabling accurate real-time predictions with minimal computational complexity during actual operation.
3Productivity
If machine learning models are trained with historical feedback data to predict quality of experience, then productivity of QoE assessment is improved, but loss of information occurs during data processing
Solution Approach 1:
The system implements a selective data retention strategy where raw feedback data is initially stored in a database, then processed through machine learning models that extract key patterns and relationships. The original detailed feedback data can be discarded after training, but the learned models retain the essential information needed for efficient QoE prediction, recovering the most valuable insights while reducing data storage requirements.
Data Source
AI summary
Processing quality of service data to provide a prediction of quality of experience is disclosed. States for a plurality of parameters affecting a level of service being provided to a set of subscribers are varied. While varying the states of the plurality of parameters, feedback from the set of subscribers is obtained indicating a level of satisfaction with the service being provided to the set of subscribers. The feedback is statistically analyzed to identifying a correlation between the states for the plurality of parameters and the level of satisfaction with the service being provided to the set of subscribers. Scores representing a predicted quality of experience for the subscribers based upon the statistically analyzed feedback are produced.