| The Internet of Things(Io T)has become one of the most dynamic and promising development areas in recent years,both technically and commercially.And with the rapid development of embedded devices and 5G technology,mobile sensing,an important sensing method in the Io T,is also evolving rapidly.Today,in the era of Artificial Intelligence(AI),the perception of the Io T as a direct reflection of real-world information provides a reliable data foundation for AI applications.However,the construction of Io T systems is limited to specific fields in most cases.The heterogeneity of devices,data and transmission protocols of different systems makes it difficult to share Io T services across domains,platforms or applications.Moreover,perceived information cannot be comprehensively utilized.As a consequence,it is difficult to form actionable knowledge for upper-layer applications to perform decisions.And due to the high spatial and temporal dynamics of mobile sensing,the heterogeneity is more complicated in the Io T systems about this.In this background,this study focuses on the status of information heterogeneity in the Io T,especially mobile sensing.By using Semantic Web technology as a means of information annotation,and taking the information description method of the mobile sensing Io T system as the research object,there are two key problems supposed to solve: registration and discovery of mobile-sensing nodes,and description modeling of mobile sensing observations.The main work of this study is as follows:(1)Studying the method of registering and discovering mobile-sensing nodes in the Io T system.In order to eliminate the heterogeneity in the mobile sensing,a framework for mobile-oriented semantic web of things,SWo MT,is proposed.In this framework,the devices in the sensing layer are abstracted as resources and provide the RESTful interfaces.And the event-driven model based on the publish/subscribe model is established,for achieving the access to mobile-sensing nodes in the Io T system by using the general Web method.Without being bound by the communication protocols in their respective systems,the heterogeneity of the system is eliminated to a certain extent.(2)Studying the semantic description methods of mobile-sensing nodes and their observations.Build the Semantic Middle Platform on the top of the resource-based sensing layer.It is used to store and manage the semantic description provided by the Io T infrastructure,providing users and developers with a unified interface to interact with the real world.Existing semantic Io T domain ontology and applications have been surveyed,and on this basis,based on the data description characteristics of mobile sensing,SWo MT ontology clusters are constructed as a semantic model to standardize information description.It provides semantic support for other modules in the Semantic Middle Platform,further eliminates heterogeneity at the semantic level.And this lays the foundation for future related research on the comprehensive application of mobile sensing data.(3)Design and implement a Mobile-oriented Semantic Web of Things Prototype System.The sensing nodes are built based on Raspberry Pi,Arduino,and Android devices.And based on the above research content,the device registration and data upload are realized.They are processed and forwarded to the application background based on the Flask framework through the Semantic Middle Platform,which is implemented by Apache Jena.Users can query and manage data in the system through Web pages and client applications.It verifies the effectiveness of the method proposed in this study.And it also has a certain value of research and application. |
PDF Full Text Request