分布式网络中 CKNN 查询的 BORA 优化汇聚

针对时空数据库中,移动对象轨迹的连续K近邻查询(continuous K nearest neighbor query,CKNN)的查询效率较低的问题,以及在分布式的移动对象数据库(moving objects databases,MOD)环境下,提升对应查询结果的数据汇聚效率问题进行了研究.在CKNN查询中,设计优化了查询海滩线的更新算法,通过在轨迹数据结构中增加更新标志位,减少了轨迹线段参与的判定运算;同时在假设的类网格覆盖的分布式空间环境下,利用基于Bresenham覆盖的路由汇聚(Bresenham-based overlay for routing and aggregation,BORA)方法,进行查询结果的汇聚;并针对不同近邻参数、轨迹数目、移动对象速度、汇聚方式等对查询时间的影响进行了仿真实验;仿真结果表明,不同参数数值的增加延长了处理时间,基于BORA的汇聚方式比一般的汇聚方式节省了更多的处理时间,提高了系统查询及处理的效率.

BORA optimized aggregation of CKNN query in distributed settings

In spatial and spatio-temporal databases and distributed settings of MOD(moving objects databases), the problem of lack of efficiency of continuous K nearest neighbor query about moving object trajectories, and that of query optimization and results aggregation were researched. The K Nearest Neighbor query efficiency is promoted by adding the update flag in trajectory data structure. And in assumed grid like coverage of spatial universe of discourse, the aggregation problem of continuous K nearest neighbor is resolved by revised BORA ( Bresenham-based overlay for routing and aggregation) algorithm. The influence of different parameters, such as K, number of trajectories, moving velocity and aggregation method were demonstrated in experiments. It can be seen that the increase of different parameter prolongs the processing time, but the BORA based aggregation saved more processing time and improved the efficiency of system.