المفاهيم الأساسية
Ensuring a seamless and positive user experience is essential for attracting and retaining customers in the telecommunication industry. This review article examines recent research and initiatives aimed at creating frameworks for assessing user Quality of Experience (QoE), and explores the integration of machine learning algorithms to enhance these tools for future advancements.
الملخص
This review article provides a comprehensive overview of the key concepts and recent developments in the field of Quality of Experience (QoE) assessment for telecommunication networks.
The article begins by defining QoE and highlighting its importance in the telecommunication industry. It outlines the four main impact factors that influence user QoE: human-related, system-related, context-related, and content-related factors.
The article then delves into the various QoE measurement tools, both closed-source and open-source, and their applications, data collection methods, and unique functionalities. Closed-source tools, developed by private companies, often provide comprehensive and user-friendly solutions with proprietary support, while open-source tools, developed through community collaboration, offer transparency, flexibility, and customization options. The review examines the strengths and weaknesses of each type of tool and their role in optimizing communication networks.
Furthermore, the article explores the integration of machine learning algorithms as a complement to these QoE measurement tools. It provides a classification of the various AI algorithms utilized across different facets of QoE management, including DNN, CNN, RFR, SVR, DT, RNN, LSTM, TCN, RBFN, GAN, and K-Means clustering. The key ideas and predictive results of these algorithms in QoE assessment are summarized.
Finally, the article addresses the existing challenges in the QoE measurement field, such as data collection, generalizability, user device diversity, interpretability, benchmarking, and privacy considerations, and envisions future directions for the development of these measurement tools.
الإحصائيات
"The jitter should be maintained at a level lower than 30 milliseconds."
"The packet loss should be kept below one percent."
"The proposed DNN-based model achieves an accuracy of 97.8% in predicting the subjective QoE scores."
"The RNN-LSTM model demonstrates the highest accuracy in predicting QoE scores."
"The proposed CNN-QoE model achieves an accuracy of 96.4% in continuously predicting QoE in streaming services."
"The RBFN-based QoE estimation model for video streaming achieves high accuracy in predicting QoE scores."
اقتباسات
"QoE is a subjective metric that incorporates human parameters and considers the user's perception, expectations, and experience along with application and network performance, providing a more comprehensive understanding of quality as experienced by end users."
"QoE represents a complete end-to-end experience. It initiates when a user sends the first request in the application to the server or acts to start up the system, while QoS solely focuses on the network and is calculated based on the delivery of packets from the server until they reach the client-side application."
"Gathering authentic user data to train AI models for precise QoE prediction is pivotal for optimizing network resources."