Modelling user experience of adaptive streaming video over fixed capacity links

Streaming video continues to experience unprecedented growth. This underscores the need identify user-centric performance measures and models that will allow operators satisfy requirements for cost-effective network dimensioning delivered with an acceptable level of user experience. paper presents analysis two novel metrics in context fixed capacity links: (i) average proportion a video’s playing time during which quality is reduced (ii) videos at least once their time, based on M/M/∞ system. Our shown hold more general M/G/∞ system metric (i), but not simulation studies show unexpected form sensitivity flow duration distribution, contrary norm increasing variance causing worse performance. At typical operational loads these new provide sensitive information rich guide understanding how degrades, than widely used throughput does. We further only combined use this existing our can holistic perspective overall