An Efficient Index for Moving Objects with Frequent Updates

With the rapid development of wireless communications and positioning systems, tracking the positions of moving objects is becoming more and more important. The TPR-tree and its variants are the dominant choices for indexing the current and anticipated near-future positions of moving objects, but the TPR-tree exhibits lower performance in the condition of frequent updates. In this paper, we propose a novel TPR-tree based indexing technique, called the TPROM- tree that could decrease the cost of frequent updates. The TPROM-tree processes updates in a memory-based way that eliminates the necessity to delete the old data item from the index during an update. Performance experiments indicate that the memory-based strategy outperforms the top-down technique used in TPR-tree and leads to indices with better query performance.