Dundar, B.Astekin, MerveAktas, M. S.2017-07-082017-07-082016978-1-5090-4795-6http://hdl.handle.net/10679/5426https://doi.org/10.1109/SKG.2016.017Due to copyright restrictions, the access to the full text of this article is only available via subscription.In this study, we introduce a big data processing framework that provides self-healing capability in the Internet of Things domain. We discuss the high-level architecture of this framework and its prototype implementation. To identify faulty conditions, we utilize a complex-event processing technique by applying a rule-based pattern-detection algorithm on the events generated real-time. For events, we use a descriptor metadata of the measurements (such as CPU usage, memory usage, bandwidth usage) taken from Internet of Things devices. To understand the usability and effectiveness of the proposed architecture, we test the prototype implementation for performance and scalability under increasing incoming message rates. The results are promising, because its processing overhead is negligible.engrestrictedAccessconferenceObject626800040170980000910.1109/SKG.2016.017Big dataInternet of thingsSelf-healing systemsPredictive maintenanceComplex event processing2-s2.0-85013249600