The Research QMon-tree Of Moving Object Indexing In Traffic Networks

With the popularity of the 4G mobile and arrival of the era of big data which based on the location service function, resulting in a large number of positioning function of the portable wireless terminal, these revolutionary changes made people travel more and more convenient. Moving MOD is the mobile object to send the current location information through the wireless device active or passive database(objects database, MOD). It is to provide the interface to the user to query the history, present and future location information of the mobile object. MOD can be used in civil aviation control, traffic management, military command, location-based information services, and other fields. At present, the most mature index model in the market for the mobile object index in the traffic network is used to build the index tree, and the index of the road network often can not meet the efficiency of the query. Therefore, it is more practical to study the index of moving objects in real life.This paper makes a deep analysis on the existing object indexing model based on traffic network in mobile of typical traffic network model and moving object indexing technique, considering the irregularity of the actual road network at the same time, makes full use of quadtree structure characteristics and puts forward an improved oriented road network mobile object index structure QMon-tree(Quad-Moving objects in networks tree). QMon-tree structure: the upper layer for an improved four tree grid spatial index structure, plus a linked list, the middle is a layer of 2DR-tree, the bottom part of the hash by moving objects. Four fork tree mechanism improved to upper area of flat space index of the whole road network is introduced, the clustering algorithm based on density road network space is divided into multiple sub index space, and set up the corresponding threshold for each of the four binary tree leaf node; information storage list for the real road, each section respectively at the lower 2DR-tree forest; middle node 2DR-tree according to the time sequence stored location information of moving object trajectories, the layer is mainly responsible for moving objects in the history and the present information index; moving object hash is composed of hash table structure and dynamic way circular linked list, mainly responsible for the location information indexing of moving objects in the future. The discrete data obtained by the method is continuous, and the accuracy of location prediction is improved by using the three subsection algorithm.Through the simulation and experiment results show that QMon-tree index structure can effectively reduce the height of the quad tree, so as to improve the road network search speed, makes the efficiency of queries become more efficient. At the same time, this paper, through piecewise cubic fitting future position trajectory, improves the accuracy of the prediction of trajectories of moving objects.