A QoE adaptive management system for high definition video streaming over wireless networks

• Published in: Telecommunication systems Vol. 77; no. 1; pp. 63 - 81
• Main Authors: Taha, Miran, Canovas, Alejandro, Lloret, Jaime, Ali, Aree
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.05.2021; Springer Nature B.V
• Subjects: Adaptive systems; Algorithms; Artificial Intelligence; Business and Management; Channel capacity; Computer Communication Networks; Digital media; Electronic devices; High definition; IT in Business; Machine learning; Multimedia; Packets (communication); Parameters; Probability Theory and Stochastic Processes; Telecommunications systems; Video data; Video transmission; Wireless communications; Wireless networks; Adaptive streaming; Wireless network; Smart algorithm; QoE assessment and management; Prediction model
• DOI: 10.1007/s11235-020-00741-2
• ISSN: 1018-4864

The development of the smart devices had led to demanding high-quality streaming videos over wireless communications. In Multimedia technology, the Ultra-High Definition (UHD) video quality has an important role due to the smart devices that are capable of capturing and processing high-quality video content. Since delivery of the high-quality video stream over the wireless networks adds challenges to the end-users, the network behaviors ‘factors such as delay of arriving packets, delay variation between packets, and packet loss, are impacted on the Quality of Experience (QoE). Moreover, the characteristics of the video and the devices are other impacts, which influenced by the QoE. In this research work, the influence of the involved parameters is studied based on characteristics of the video, wireless channel capacity, and receivers’ aspects, which collapse the QoE. Then, the impact of the aforementioned parameters on both subjective and objective QoE is studied. A smart algorithm for video stream services is proposed to optimize assessing and managing the QoE of clients (end-users). The proposed algorithm includes two approaches: first, using the machine-learning model to predict QoE. Second, according to the QoE prediction, the algorithm manages the video quality of the end-users by offering better video quality. As a result, the proposed algorithm which based on the least absolute shrinkage and selection operator (LASSO) regression is outperformed previously proposed methods for predicting and managing QoE of streaming video over wireless networks.