A Systematic Examination of Inter-App Conflicts Detections in Open IoT Systems

Recent several years have witnessed the rapid development of open platforms for developing and deploying Internet of Things (IoT) apps. One of the important issues in such environments is the unintended interactions among independently developed apps installed in a single environment. Although there have been some recent studies trying to address the issue, the treatments to the problem remain preliminary, reflected by the rudimentary definitions of inter-app conflicts and the entailed inadequacy of the previously proposed solutions. This paper strives to provide a deeper and more systematic investigation to the problem. It offers a series of new definitions and categorizations to more precisely characterize the nature of IoT inter-app conflicts, and proposes a new representation named IA Nets for representing IoT controls and inter-app interplays. It then describes an efficient conflicts detection algorithm developed upon these new definitions and representations. It further presents DIAmond, a compiler and runtime framework that integrates all the proposed techniques together into a comprehensive solution. Experiments on SmartThings apps validate its effectiveness for IoT inter-app conflicts detections.