![]() ![]() Thus, it is not representative for mobile usage. The desktop version used in modern web browsers behaves differently from Android or iOS versions, as different libraries, application types, and operating systems come into play 13, 16, 17. The available archives mainly concentrate exclusively on data collection via the desktop version of YouTube 13, 14, 15. But in literature are only a few datasets available that document the use of YouTube in mobile environments. In fact, YouTube generates over a fifth (20.4% downlink, 5.4% upstream) of all global mobile Internet traffic 12. This makes YouTube the second most visited website in the world 6, 7, with mobile access accounting for two-thirds of the platform’s video views in the second quarter of 2021 8, 9, 10, 11. Statistically speaking, every person in the world watches an average of 7.55 minutes on YouTube every day 5. Overall, approximately 1.24 billion monthly users watch nearly 1 billion hours of online video on YouTube every day 3, 4, 5. Thus, the study of current mobile traffic and prediction of future traffic for accurate network management is a challenging task without a data basis. However, datasets on Internet traffic dealing with mobile streaming of videos from major streaming platforms are currently sparsely treated in literature. With more than 4.9 billion Internet users worldwide 1, 2, online videos and entertainment are among the most popular activities of users, and thus, of particular interest. The nationwide rollout of new mobile communication technologies requires in-depth traffic analysis, usage studies, and network management. We present the most relevant quality indicators for scientific use, i.e., initial playback delay, streaming video quality, adaptive video quality changes, video rebuffering events, and streaming phases. This corresponds to 332 GB video payload. The dataset includes 80 network scenarios with 171 different individual bandwidth settings measured in 5,181 runs with limited bandwidth, 1,939 runs with emulated 3 G/4 G traces, and 4,022 runs with pre-defined bandwidth changes. With this public dataset, we provide 1,081 hours of time-synchronous video measurements at network, transport, and application layer with the native YouTube streaming client on mobile devices. Thus, in view of the continuous technological advances in end devices and increasing mobile use, datasets for mobile streaming are indispensable in research but only sparsely dealt with in literature so far. ![]() According to Google statistics, three out of five video views come from mobile devices. As a result, streaming is by far the predominant type of traffic in communication networks. Around 4.9 billion Internet users worldwide watch billions of hours of online video every day.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |