Publication detail

On Predicting Video Quality Expectations of Mobile Users

HOŠEK, J. UHLÍŘ, D. KOVÁČ, D. GALININA, O. ANDREEV, S. KOUCHERYAVY, Y. RIES, M.

Original Title

On Predicting Video Quality Expectations of Mobile Users

Czech Title

Předpovídání spokojenosti mobilních uživatelů při používání video služby

English Title

On Predicting Video Quality Expectations of Mobile Users

Type

conference paper

Language

en

Original Abstract

Mobile network operators are currently seeking for simple but accurate methods to predict the levels of satisfaction for their customers using the on-line multimedia applications, such as YouTube. Even though the ultimate user demands are known to be influenced by multiple factors, there is one clear trend – people require an increasingly higher quality of mobile video services. To this end, modeling the corresponding quality of experience (QoE) constitutes a non-trivial task and calls for a careful balance between the key underlying aspects, while maintaining the overall complexity as low as possible. This should in turn deliver the much needed usability of the resulting model across many real-world scenarios, and in this work we develop a novel QoE prediction model based on our extensive user experience investigation of the YouTube service. Our proposed solution allows network operators to estimate the degrees of video quality and thus predict the associated mobile user expectations in their deployments. The design principles behind our methodology, its accuracy evaluation, as well as the obtained numerical results are reported in the course of this paper.

Czech abstract

Operátoři mobilních sítí v současné době hledají jednoduché, ale přesné metody, jak předvídat úroveň spokojenosti svých zákazníků používající on-line multimediální aplikace, jako je například YouTube. I přesto, že konečné požadavky uživatele jsou vždy ovlivněny mnoha faktory, existuje jeden jasný trend - lidé vyžadují stále vyšší kvalitu mobilních video služeb. Z tohoto důvodu představuje modelování kvality zážitku (QoE) netriviální úkol a vyžaduje pečlivou analýzu všech klíčových aspektů avšak s ohledem na zachování celkové složitosti predikčního algoritmu tak nízké, jak je to jen možné. To by mělo na oplátku přinést tolik potřebnou použitelnost výsledného modelu v mnoha reálných scénářích. S ohledem na tyto předpoklady jsme v rámci této práce vyvinuli nový QoE predikční model založený na naší rozsáhlé studii uživatelské spokojenosti se službou YouTube. Náše řešení umožňuje síťovým operátorům odhadovat úrovně kvality videa, a tak předpovídat odpovídající subjektivní hodnocení mobilních uživatelů. Principy návrhu stojící za použitou metodikou, hodnocení její přesnosti, jakož i získané numerické výsledky jsou uvedeny v tomto článku.

English abstract

Mobile network operators are currently seeking for simple but accurate methods to predict the levels of satisfaction for their customers using the on-line multimedia applications, such as YouTube. Even though the ultimate user demands are known to be influenced by multiple factors, there is one clear trend – people require an increasingly higher quality of mobile video services. To this end, modeling the corresponding quality of experience (QoE) constitutes a non-trivial task and calls for a careful balance between the key underlying aspects, while maintaining the overall complexity as low as possible. This should in turn deliver the much needed usability of the resulting model across many real-world scenarios, and in this work we develop a novel QoE prediction model based on our extensive user experience investigation of the YouTube service. Our proposed solution allows network operators to estimate the degrees of video quality and thus predict the associated mobile user expectations in their deployments. The design principles behind our methodology, its accuracy evaluation, as well as the obtained numerical results are reported in the course of this paper.

Keywords

Anylitcial modelling;Mobile network;QoE;Subjective evaluation;YouTube

RIV year

2015

Released

08.10.2015

Publisher

IEEE Xplore

Location

Brno, Czech Republic

ISBN

978-1-4673-9282-2

Book

2015 7th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT)

Edition

1

Edition number

1

Pages from

110

Pages to

115

Pages count

6

BibTex


@inproceedings{BUT118090,
  author="Jiří {Hošek} and Dalibor {Uhlíř} and Dominik {Kováč} and Olga {Galinina} and Sergey {Andreev} and Yevgeni {Koucheryavy} and Michal {Ries}",
  title="On Predicting Video Quality Expectations of Mobile Users",
  annote="Mobile network operators are currently seeking for simple but accurate methods to predict the levels of satisfaction for their customers using the on-line multimedia applications, such as YouTube. Even though the ultimate user demands are known to be influenced by multiple factors, there is one clear trend – people require an increasingly higher quality of mobile video services. To this end, modeling the corresponding quality of experience (QoE) constitutes a non-trivial task and calls for a careful balance between the key underlying aspects, while maintaining the overall complexity as low as possible. This should in turn deliver the much needed usability of the resulting model across many real-world scenarios, and in this work we develop a novel QoE prediction model based on our extensive user experience investigation of the YouTube service. Our proposed solution allows network operators to estimate the degrees of video quality and thus predict the associated mobile user expectations in their deployments. The design principles behind our methodology, its accuracy evaluation, as well as the obtained numerical results are reported in the course of this paper.",
  address="IEEE Xplore",
  booktitle="2015 7th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops (ICUMT)",
  chapter="118090",
  edition="1",
  howpublished="electronic, physical medium",
  institution="IEEE Xplore",
  year="2015",
  month="october",
  pages="110--115",
  publisher="IEEE Xplore",
  type="conference paper"
}