An Efficient Network-Based QoE Assessment Framework for Multimedia Networks Using a Machine Learning Approach
The Internet is integral to modern life, influencing communication, business, and lifestyles worldwide. As dependence on Internet services grows, so does the demand for high-quality service delivery. Service providers must uphold high standards of quality of service and Quality of Experience (QoE) t...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Open Journal of the Communications Society |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10892313/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The Internet is integral to modern life, influencing communication, business, and lifestyles worldwide. As dependence on Internet services grows, so does the demand for high-quality service delivery. Service providers must uphold high standards of quality of service and Quality of Experience (QoE) to ensure user satisfaction. QoE, a key metric for multimedia services, reflects user satisfaction with service quality. However, measuring QoE is challenging due to its subjective nature and the complexities associated with real-time feedback.This paper presents an open-source framework for assessing QoE in multimedia networks using only key network parameters. By eliminating the need for video-specific data, this framework simplifies the traditional ITU standard for QoE assessment, achieving high accuracy in predicting Mean Opinion Scores (MOS). The framework leverages Machine Learning (ML) to model the relationship between network parameters and QoE, providing a scalable and efficient solution for real-time QoE evaluation in multimedia networks.By focusing exclusively on network metrics (e.g., delay, jitter, and packet loss), it eliminates the need for video-specific parameters to calculate MOS. Addressing the limitations of existing QoE models, the framework integrates real-time data collection, ML predictions, and adherence to international standards. This reduced-parameter approach achieves approximately 97% of the prediction accuracy of the full ITU P.1203 implementation while significantly lowering data requirements and computational demands. By enabling ITU-T P.1203 MOS score calculation without video-specific data, the framework offers a faster, scalable solution adaptable to diverse real-time multimedia services. |
|---|---|
| ISSN: | 2644-125X |