A fast and compact indexing technique for moving objects

Abstract

Advances in ubiquitous connectivity and location sensing technology have fuelled a rich collection of location based services (LBSs). Efficient spatial indexing techniques are one of the most effective optimization methods to improve the quality of services. Although a variety of spatial index structures like R-tree family and grid variant index structures have been proposed and deployed in real time location based service provisioning systems, they are known to perform poorly when there is high degree of the skewedness in both density distribution and spatial resolution of mobile objects. First, it is hard to decide the optimal resolution of the grid structure. Second, it is equally hard to build a balanced R-Tree like index structure that is effective in handling highly skewed distribution of mobile objects. With these issues in mind, we introduce the concept of spatial order sequences and propose a fast and compact index structure for moving objects by utilizing spatial order sequences through a number of density-conscious optimizations. First, we propose the concept of Ordered-Cell Group (OCG) and design a OCG based grid index structure. Second, we speed up the search efficiency of OCGs by effective compaction of identifiers of OCG cells to maximize the fan-out of index node and decrease the depth of the index. Finally, we develop an efficient query processing algorithm that can effectively utilize OCG cells to speed up the processing of spatial queries. Our experimental results demonstrate the effectiveness of our approach compared to existing index techniques.