A review on intelligent IoT systems design methodologies

Abstract The design of distributed real-time systems including In- ternet of Things (IoT) systems is complex due to open-ended architectures requirements. An emergent solution to deal with this complexity and to manage system constraints is the integration of adaptation strategies in distributed devices behavior. The developement of such systems denotes a signifi-cant challenge on providing autonomous and intelligent ability to verify distributed devices behaviors and constraints such as battery storage of sensors. Existing AI-based approaches for dynamic IoT systems proposed solutions relative to the target system for resolving specific problems without consid-ering interoperability and reusability. Using MDE approach and UML/MARTE profile for high-level abstraction becomes a promising solution to ease the design of intelligent IoT systems. This paper recalls and classifies the existing AI and MDE-based works for the design of dynamic distributed systems especialy IoT applications. We concentrate on a set of criteria to highlight the shortages of existing approaches on the design of distributed devices interaction, adaptation and system constraints. Finally, we introduce our future directions to cope with the limits of existing solutions while taking into account the observed criteria.