摘要
传统的缓存置换策略未充分考虑数据访问的空间特征,也不适用于基于矢量瓦片的替换。该文根据矢量瓦片的空间数据结构,提出一种适用于矢量瓦片缓存替换的视点相关预测区域算法:首先根据瓦片存储中多分辨率金字塔结构进行空间单元划分,并根据用户操作类型求解矢量瓦片及空间单元热度,从而构建用户视点位置相关的预测区域;然后综合考虑瓦片层级、空间单元热度及距离等因素进行预测区域分析,获得瓦片缓存价值并进行瓦片置换。通过与传统的FIFO、LRU和LFU缓存策略相比较,该算法的瓦片命中率比FIFO和LRU分别提高了近50%和20%,瓦片的请求耗时分别缩短了50%和30%左右,相比LFU也有明显优势。该研究为WebGIS提供了一种更具潜力的瓦片缓存方法。
The traditional cache replacement strategy doesn't fully consider the spatial characteristics of data access,so it's not suitable for vector tile.According to the spatial data structure of vector tiles,a predicted region algorithm related to viewpoint for vector tile cache replacement is proposed in this paper.The algorithm divides space units based on the characteristics of multi-resolution pyramid structure.Combined with the users'operation types,the vector tile heat and space unit heat are calcu-lated to construct the predicted region.And then considering the tile level,the space unit heat and the distance factor,etc.,the predicted region analysis is finally carried out to obtain the tile cache value and perform tile replacement.Compared with tradi-tional caching strategies such as FIFO,LRU and LFU,the results show that the tile hit rate of the proposed algorithm is 50%and 20%higher than that of FIFO and LRU,and the tile request time is reduced by 50%and 30%respectively compared with FIFO and LRU.Compared with LFU,this algorithm also has obvious advantages and is suitable for fast visualization of vector tiles.The study provides a potential tile cache method for WebGIS.
作者
王治铭
范光鹏
陈飞翔
崔晓晖
WANG Zhi-ming;FAN Guang-peng;CHEN Fei-xiang;CUI Xiao-hui(School of Information Science and Technology,Beijing Forestry University,Beijing 100083;Engineering Research Center for Forestry-Oriented Intelligent Information Processing of National Forestry and Grassland Ad ministration,Beijing 100083,China)
出处
《地理与地理信息科学》
CSCD
北大核心
2021年第1期5-10,46,共7页
Geography and Geo-Information Science
基金
国家重点研发计划项目(2017YFD0600906)。
作者简介
王治铭(1996-),男,硕士研究生,研究方向为空间信息技术、地理信息系统。;通讯作者:崔晓晖,E-mail:cuixiaohui@bjfu.edu.cn。
