多用户连续k近邻查询多线程处理技术研究

Research on multi-threading processing of concurrent multiple continuous k-nearest neighbor queries

针对面向移动对象集的多用户连续k近邻查询处理,提出了基于多线程的多用户连续查询处理(MPMCQ)框架,采用流水线处理策略,将连续查询处理过程分解为可同时作业的查询预处理、查询执行以及查询结果分发三个执行阶段,利用多线程技术来提高多用户连续查询处理的并行性;基于MPMCQ框架和移动对象内存格网索引,提出了基于多线程的连续k近邻查询处理(MCkNN)算法。实验结果与分析表明,基于MPMCQ框架的MCkNN算法在多核平台上优于CPM、YPK-CNN等现有算法。

To deal with the multiple concurrent continuous k nearest neighbors queries towards moving objects, the proposed a Multi-threading Processing of Multiple Continuous Queries （MPMCQ） framework, which adopted pipeline strategy and departed the continuous query processing into three simultaneous stages： query processing, query executing and query results dispatching to improve the parallelism with muhi-threading technology. Based on MPMCQ framework and grid index for moving objects, a Multi-threading processing of Continuous k Nearest Neighbor queries （MCkNN） algorithm was also presented. Experimental results and analysis show that MCkNN algorithm outperforms CPM, YPK-CNN, etc.