Improving failure prediction accuracy in smart environments

S. Ozen, E.U. Warriach, T. Özcelebi, J.J. Lukkien

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

2 Citations (Scopus)
5 Downloads (Pure)


Smart Environments (SE) and Internet of Things (IoT) are two concepts that connect consumer electronics (CE) to each other and to the Internet domain. This enables various applications where CE devices work together to achieve goals of their users by communicating over a network. Application failures due to fluctuation of resources and environmental factors must be prevented, even though it is a challenge in such a distributed platform. Application adaptation is used extensively for this purpose. Since triggering of unnecessary adaptations affects application quality of service and user experience negatively, we propose a failure prediction method that makes a distinction between potential failures and outliers. The experiments show that the number of unnecessary adaptations is reduced significantly by the proposed method, while maintaining dependability.
Original languageEnglish
Title of host publicationProceedings of the 2016 IEEE International Conference on Consumer Electronics (ICCE), January 7-11, 2016, Las Vegas, NV
EditorsF.J. Bellido, N.C.H. Vun, C. Dolar, D. Diaz-Sanchez, W.-K. Ling
Place of PublicationPiscataway
PublisherInstitute of Electrical and Electronics Engineers
ISBN (Print)978-1-4763-8363-9
Publication statusPublished - 2016
Event2016 IEEE International Conference on Consumer Electronics (ICCE 2016) - Las Vegas, United States
Duration: 7 Jan 201611 Jan 2016


Conference2016 IEEE International Conference on Consumer Electronics (ICCE 2016)
Abbreviated titleICCE 2016
Country/TerritoryUnited States
CityLas Vegas


Dive into the research topics of 'Improving failure prediction accuracy in smart environments'. Together they form a unique fingerprint.

Cite this