面向多图层的空间矢量数据缓存更新策略研究  

Research on cache replacement strategy for space vector data of multi-layer

作　　者：吴家皋[1] 张曦[1] 庄嵩杰[1] 邹志强[1] 胡斌[2] 

机构地区：[1]南京邮电大学计算机学院计算机技术研究所,南京210003 [2]南京师范大学虚拟地理环境教育部重点实验室,南京210046

出　　处：《计算机工程与应用》2014年第7期89-93,共5页Computer Engineering and Applications

基　　金：国家高技术研究发展计划(863)(No.2009AA12Z219);国家自然科学基金(No.40801149);东南大学计算机网络和信息集成教育部重点实验室开放研究基金(No.K93-9-2010-06)

摘　　要：为提高P2P空间矢量数据索引网络的性能,在已有混合结构P2P空间索引网络的基础上,引入缓存机制,并提出了一种新的面向多图层的空间矢量数据缓存更新策略。该策略针对空间矢量数据多图层特性,综合考虑图层优先级以及查询频率对于缓存更新的影响,合理地利用了缓存空间。同时,将缓存更新抽象成0/1背包问题的数学模型,采用遗传算法对其优化求解。仿真结果表明该缓存更新策略可以增加缓存命中率,提高空间索引效率。In order to promote the performance of P2P spatial indexing network on vector data,a caching mechanism is introduced into the existing hybrid P2P spatial indexing network.And a novel cache replacement strategy for space vector data of multi-layer is proposed.In allusion to the multi-layer characteristics of the spatial vector data,the layer priority and query frequency are considered,which make full use of the cache space.Furthermore,the cache replacement is abstracted as a mathematical model of the 0/1 knapsack problem and solved by genetic algorithms.The simulation experiments indi-cate that this strategy can increase the caching hit ratio and improve the indexing efficiency effectively.

关 键 词：索引网络 缓存更新 多图层 0 1背包问题 遗传算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]