Research On Index And Query Technologies Of Resource For Cyber-physical System

With embedded computing, sensor monitoring, wireless communication technology and large-scale data storage processing technology, through the development of a sense of control ability, communication, numeracy, and information is the process of combining with the depth of the physical process, resulting the set of computing, communication and control as one integration called Cyber-Physical System(CPS). CPS has wide application background, it can be used for intellingent transportation, smart cities, national defense, environmental monitoring, health care and other fields, since the concept of CPS has been proposed, domestic and foreign academics make a in-depth and extensive discussion on the relevant challenges and applications. This paper studies the indexes and related queries issues in the CPS data resource management.Multidimensional, dynamic and spatial-temporal peculiarity of resources is the main characteristics of the Cyber-physical system. And from physical domain to information domain, it makes the data which is corresponding to resources must be multidimensional, scalability and strong spatial-temporal charaterictics in the process of storage and organization. In the CPS resource management, how to make it rational and effective for data storage, organization and management, so that when a user queries it can quickly and efficiently provide accurate data is an important issue in the field of CPS research. R-tree index structure and its evolution has good scalability, strong spatial-temporal resistance and the ability to handle high-dimensional data, so it has certain advantages in the data organization and management.Thesis based on the analysis of domestic and foreign CPS resource management, index structures and query optimization algorithms, do some research on the CPS static and dynamic resources indexing and query respectively. Firstly, for the static resources, adding resources capacity indicator based on the IR-tree, we proposed a new index structure named AIR-tree which includes location-based, text-matching and capabilities. And we use a road network model for the calculation of the distance and prosed a relevant top-k query algorithm. Finally we make a comparative experiment with a road network travesral query algorithm that proved the resource query more efficient based on this index. Then for the moving objects trajectories issue in dynamic resource, considering the location information and text description of the trajectory, combined the inverted file and STR-tree together to build a new index structure named STIR-tree, and prosed the trajectory k nearest neighbor query algorithm(T-k NN). Mainly solved the trajectory modeling, index building, filtering strategy of the query algorithm and distance computing between trajectories and some other difficult problems. At last, verified by simulation experiment the rationality of the STIR-tree index structure and the validity of the T-kNN query algorithm.