Research And Application Of Entity Description And Search Technology In Internet Of Things

The IoT search service is one of the most important services provided by the Internet of Things.However,how to efficiently and accurately obtain the information that best meets user needs from massive,heterogeneous and dynamic IoT entity data is a major challenge currently facing.First of all,there is no unified description model for the IoT entity resources,which leads to the heterogeneity of the resource description and brings difficulties to the data collection and processing of the IoT entity search system.Secondly,the access of large-scale heterogeneous IoT devices makes the IoT entity data have characteristics such as mass,polymorphism and relevance.Traditional Web search technologies are no longer applicable to the IoT environment.Therefore,this thesis studies the IoT entity description and search technology,and implements an agricultural Internet of Things entity search prototype system to verify the feasibility of the proposed method.The main research contents of this thesis are as follows:(1)Aiming at the problem of lacking a unified IoT entity resource description model,since ontology can eliminate conceptual ambiguity,this thesis proposes an ontology-based entity description model,IoTEDM,which describes entities uniformly from six aspects: the entity's natural attributes,status attributes,behavior operations,event notifications,history records,and security requirements,and instantiates the entity using JSON format.(2)Aiming at the massive,dynamic and non-uniform search space of IoT entities,clustering algorithms can be used to reduce the scope of entity search.However,the non-uniformity of IoT entity deployment makes the classic k-means and DBSCAN algorithms unable to divide the appropriate entity clusters.Therefore,this thesis proposes an advanced density clustering algorithm,A-DBSCAN,which first performs density clustering based on location,then performs k-means secondary division on large-scale entity clusters,and processes noise points.Simulation results show that in the case of uneven distribution of entities,the entity search method based on A-DBSCAN greatly reduces the scope of entity search and improves overall search efficiency.(3)Based on the IoTEDM,an agricultural IoT entity search prototype system based on ADBSCAN was designed and implemented,which verified the feasibility of the proposed method.