面向轨迹数据查询优化的缓存机制研究

Research on Cache Mechanism for Query Optimization of Trajectory Data

移动对象数据库(Moving object database,MOD)管理随着时间不断改变位置的空间对象,其已经在广泛的应用中被研究。尽管索引和查询算法等许多与MOD相关的技术已经被提出,但缓存管理仍然被忽略,这对于数据库性能至关重要。传统的缓存方法忽视了数据本身的时空特性,无法实现良好的性能。本文提出从缓存层面充分挖掘轨迹数据的查询性能,首先针对轨迹数据特有的存储结构和读写过程,设计了适合MOD的缓存访问机制;然后对于MOD系统中缺少与应用场景和访问模式相关的缓存策略问题,定制了缓存替换方法;最后开发了缓存管理工具MOCache,以可视化的形式动态跟踪查询语句结束后缓存状态的变化。通过实验表明本文提出的缓存替换策略与传统的策略相比,命中率提高到76.56%,缩短了查询时间,并且使用缓存工具监控历史状态信息,能够更加全面地反馈和分析性能问题。

Moving objects database(MOD)manage spatial objects that continuously change their locations over time and have been studied in a wide range of applications.Although a number of relevante techniques have been proposed such as indexing and query algorithms,cache management in MOD has been ignored.This is essentially important for database performance.Traditional cache methods ignore the spatial-temporal characteristics of data and cannot achieve good performance.This paper proposes to fully exploit the query performance of trajectory data from the cache level.Firstly,based on the unique storage structure and read/write process of trajectory data,a cache access mechanism suitable for MOD is designed.Then,due to the lack of cache policies related to application scenarios and access modes in MOD,a cache replacement strategy is customized.Finally,a cache management tool MOCache for trajectory data is implemented.By utilizing MOCache,dynamic tracking of cache state changes is visualized after each query statement.Compared with traditional algorithms,the proposed cache replacement strategy improves the hit ratio to 76.56%and reduces query time,and using the cache tool to monitor historical state information can facilitate comprehensive feedback and analyze performance problems.