Challenges and Solutions for Video Streaming Services Quality of Experience (QoE) Prediction Toward 6G Wireless Networks

This paper gives a comprehensive review of the problems and developments in predicting Quality of Experience (QoE) for video streaming services with special emphasis on the future 6G wireless networks. The rapid growth of mobile communication technologies has transformed the consumption of video content, leading to an explosion in video streaming traffic, which is set to continue increasing with the international proliferation of mobile devices and internet users. Video streaming services like Netflix, YouTube, and Amazon Prime are seeing growing demand, and delivering maximum QoE has become a critical challenge for service providers. This paper highlights the growing importance of being capable of accurately predicting QoE in a heterogeneous and dynamic environment on the basis of user behavior, network conditions, device capabilities, and the nature of the content streamed. The review discusses the use of deep learning techniques, such as Long Short-Term Memory (LSTM) networks, Convolutional Neural Networks (CNN), and Support Vector Machines (SVM), graph neural networks (GNNs), which have been shown to address the challenges of real-time QoE prediction. These models are capable of handling big data and temporal relations, hence suitable for QoE prediction in video streaming contexts. Moreover, the paper also compares the performance of subjective and objective methods for QoE measurement, with a clear comparison of various deep models based on predictive performance, complexity, and usefulness. As regards 5G and 6G, the review also covers the use of adaptive bitrate algorithms and quality of service management systems that dynamically adjust the video quality based on the varying network conditions. With the evolution of such technologies, real-time scalable QoE prediction models become increasingly vital in order to provide user satisfaction and service reliability. The contributions of the paper are an extensive study of the difficulties in QoE prediction, comparison of deep learning models, and recommendations for future work towards enhancing QoE prediction for video streaming services so that they can cope with the needs of future wireless communication networks.