A multi-layer probing approach for video over 5G in vehicular scenarios

• Autor(a) principal: LOPES, RUI
• Data de Publicação: 2022
• Outros Autores: Rocha, Filipe, Sargento, Susana, Luís, Miguel, Leitão, Ricardo, Marques, Eduardo, Antunes, Bruno
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.21/15543
• topic: Video transmission; 5G; MPEG-DASH; Key performance indicators; Mobile networks

Fifth generation (5G) technologies are becoming a reality throughout the world. In parallel, vehicular networks rise their pace in terms of utilization; moreover, multimedia content transmissions are also getting an always increasing demand by their users. Besides the promised performance of 5G networks, several questions still arise among the community: are these networks capable of delivering high quality video streaming services in moving scenarios? What is the relationship between the network conditions and the video quality of experience? To answer to the previous questions, in this paper we propose a multi-layer probing approach able to assess video transmissions over 5G and 4G, combining data from all layers of a communication model, relating events from its origin layers. The probe's potential is thoroughly evaluated in two distinct video streaming use cases, both targeting a vehicular scenario supported by cellular 4G and 5G networks. Regarding the probe's performance, we show that a multitude of performance and quality indicators, from different stack layers, can be obtained. As for the performance of 4G and 5G networks in video streaming scenarios, the results have shown that the 5G links show a better overall performance in terms of video quality-of-experience, granting lower delays and jitter conditions, thus allowing video delay to be diminished and segment buffering to be better performed in comparison to 4G, while still showing adaptability in lightly traffic-saturated vehicular-to-vehicular scenarios.