Research On The On-line Analysis And Processing Technology For Massive Moving Object Data On Internet Of Things Environment

With the rapid development of the Internet of Things technology such as RFID and wireless LAN, a large number of moving objects data is produced. These moving objects generally have multi-dimensional attributes, the spatial and temporal characteristics. These moving objects communicating with other moving objects generate moving objects network. How to handle and effectively use the massive data generated by these moving objects, and apply warehouse and on-line query analysis to the moving objects network become one of the research focuses.Considering that the moving objects network contains specific very valuable substructure information, there are many structured data query algorithms, but most of these algorithms are designed for static data. However, these algorithms are very difficult to play a role in processing the moving objects data generated by the moving objects. In this article, we study the on-line analysis and processing of the structure data on the moving objects network. Another important issue is how to build a moving object graph cube generated by a large number of aggregate graphs which is unlike traditional numeric data. These aggregate graphs contain a plurality of nodes and have a complicated structure, which requires a lot of storage space. And it will take more time to process and analysis these aggregate graphs data. How to compress so many aggregate graphs data has become an urgent problem. In this paper we do some depth research on the moving objects data, efficiently compress the nodes and edges of the aggregate graphs, and provide the effective help for the users to analysis the moving objects data.The main contributions and innovations of the paper are as follows:1) In this paper, we introduce a new graph cube to handle a large number of moving objects structured data. It can effectively support OLAP queries in the moving objects on the moving object network. Taking into account the vertex attribute aggregation and an overview of the network structure, the graph cube beyonds the numerical data cube model, produces a number of meaningful structure data and rich aggregate networks.2) In this article, a new kind of OLAP query which is different from traditional cube query is proposed, called muti-boid query. This query can be used in the moving objects network. And we implement graph cubes through the combination of moving objects network characteristics and existing data cube technology. 3) Aiming at the amount of aggregation graphs of the massive moving objects network produced by building graph cube, we introduce the idea of compressing graph and put forward the MC-compress algorithm of compressing aggregation graphs, which merges vertices and edges in the aggregation graph. By comparing the largest differences of weight of two edges when merging two vertexes in the aggregation graph, we can find the optimal merging pairs of vertices. Finally, compressed graph constructed by the super vertexes and super edges is generated. This method speeds up the display of graph structure after querying aggregation graph in graph cube, and reduces the space that is used to store the large amounts of aggression graph in the process of building the graph cube.4) Aiming at the amount of aggregation graphs produced by a multidimensional moving object network in the dimension which the user interests in. We introduce the idea of graph index and put forward the MCPath algorithm of the structure data query to solve the problem of querying the interesting structure data on specific dimension. The MCPath algorithm decomposes the graph to be shortest paths, and then joins those candidate paths. Finally, the structural data which the user need to query is generated. The algorithm turns the former method of querying only one node once into the method of querying one path once, reduces the time of querying structure data, and accelerates the speed of querying structure data.