3DR-Tree Model Improvement Based on Enhance of Index Performance

3DR-tree is an index method using the traditional R-tree to index moving objects. Its defect is that a cube generates when an object remains still for a period of time. For those objects that remain still for a long period of time, many strip cubes will generate. MBR will be overlong or overlarge, thus increasing a great deal of overlap and reducing index performance greatly. This paper is to improve the 3DR-tree model to enhance index performance. On the basis of 3DR-tree, this paper will put forward improving historical data index performance of 3DR-tree through node splitting. 3DR-tree is provided with the online data index function by means of tree splitting.