Research On Key Technologies Of Real-time Semantic Processing For Sensor Data In Micro-environment Monitoring Platform

The rapid development of sensor technology offers a wide range of opportunities for real-time applications in all areas of our lives,such as environmental monitoring,healthcare,and urban traffic management.Due to the rapid emergence of these applications,the demand to analyze the data streams of the sensor is growing.The main method of enriching the sensor description is to make semantic information integrated into the data stream,transforming the low-level observation data into higher-level observation data.By applying semantic reasoning,the implicit or new knowledge can be derived from the semantically labeled sensor data,thus achieving intelligent decision making.However,the semantic annotation and reasoning of sensor data streams are still in an immature state.And the extensive attentions are required to support the more efficient annotation and reasoning.The main research work of this paper is as follows:1)Design of real-time semantic annotation method for sensor data stream.First,based on the format of the sensor data stream and the structure of the SOSA/SSN ontology,the S-SASML mapping language and SDS2 R algorithm are designed,which make the sensor data stream conform to the structure of SOSA/SSN ontology.Moreover,the thread pool technology is used to achieve high concurrent processing,thus improving the real-time performance of our method.2)Design of RDF stream processing method based on Spark Streaming.First,the window operations provided by Spark Streaming extracts the RDF stream into the window.Then,the filtering,mapping and aggregation operations of the window-based RDF stream are performed by using these operations(e.g.filter,map and reduce)provided by Spark Streaming,in order to obtain the aggregated RDF stream data.Finally,according to the application requirements,the Jena Rule are applied to the aggregated RDF data stream,and the Jena inference engine is used for rule reasoning.3)Design and implementation of the prototype system.Firstly,for the sensor data in micro-environment monitoring platform,the demands of the real-time semantic processing prototype system are analyzed,which include functional requirements and system use case.Then,according to the needs analysis,each module of the prototype system is designed.Finally,in the development environment MyEclipse,the Java language is used to develop the user interface,realizing the micro-environment monitoring,real-time analysis,online real-time display of various data.