Research And Implementation Technology On Index Of Moving Objects

Now users can easily acquire their continuously changing position, with the help of rapidly developing technology like mobile communication and Global Positioning System. Accordingly, Moving Objects Databases (MOD) recently becomes a hot issue in database related research fields. MOD stores moving objects as well as spatial information of their position, and also supports queries for their historical, current and anticipated future position. Indexes of moving objects provide a vital foundation upon which to implement MOD products.cTPR-Tree is a clustering based index for moving objects databases. It improves the split procedure of TPR-Tree to enhance the update performance. First, cTPR-Tree extends the k-means algorithm to enable it to handle moving objects efficiently. Then, the extended k-means algorithm is adopted into the split procedure. Thanks to the advantage that k-means algorithm uses the object number of clusters as an input parameter, the traditional two-way split method is upgraded to a multi-way split method. The new split procedure accommodates the distribution of moving objects so well that, the time cost of updates is reduced and the index quality is improved. Experiments show that the update performance of cTPR-Tree outperforms TPR-Tree. At the same time, the query performance is also improved.GAIP is a generalized analysis platform for indexes. It improves the tree methods of GiST, and provides a generalized framework independent of data types and algorithm strategies to implement indexes. Within this unified framework, GAIP defines performance metrics that reflect performance loss, measured in I/Os needed for an input workload. The tree which minimizes the total number of I/Os for the workload can be considered a more efficient way. The advantage of using a common metrics is that the performance of different indexes can be evaluated easily and effectively. Moreover, visualization facilities are integrated into the analysis framework, which helps designers trace the change of index and focus on the sources of poor performance.