Autonomic Coordination of IoT Device Management Platforms

With the adoption of the Internet of Things (IoT), devices become integrated in our daily lives. These devices are more and more interdependent, either by guaranteeing connectivity, or by providing data for complex services. In an ever-evolving IoT context, continuously managing these connected objects is crucial, to handle vulnerabilities, ensure their well-functioning, and add new services. These features fall under the scope of Device Management (DM). The distribution and scalability of DM platforms become all the more exacerbated as device profiles and deployment use cases get more heterogeneous. Thus, multiple actors are involved in IoT services, with their respective, isolated, DM platforms. However, DM operations, such as reboots and firmware updates, have a direct impact on the state of devices, which requires awareness across DM platforms. In this paper, we address conflicts linked to connectivity dependencies, which may arise during the simultaneous execution of DM operations on interdependent devices. Hence, we propose a coordinator that collects DM operation intentions from isolated DM platforms and detects the relevant dependencies, so as to plan and orchestrate the execution of these operations. An evaluation of this middleware is presented, with perspectives for industrial applications.