位置相关查询中基于最小访问代价的缓存替换方法

Cache replacement method based on lowest access cost for location dependent query

在位置相关查询(LDQ)中由于用户的移动性和数据的位置相关性,给缓存替换策略带来了新的挑战。在详细分析位置相关数据(LDD)的空间位置特性和几种典型的位置相关缓存替换策略的基础上,提出一种基于最小访问代价的缓存替换策略(PLAC),一些重要的缓存替换因素如访问概率、更新频率、数据距离和有效范围等都包含在代价函数里,PLAC根据代价函数值的大小来决定被替换的数据,由此来保证有限缓存的最大使用率。通过实验对比,PLAC比其他位置相关缓存替换策略更为有效地提高了缓存命中率,缩短了查询平均响应时间。

Because of the user＇s mobility and the location dependency of data, new challenge has been brought to cache replacement strategy for Location Dependent Query （ LDQ）. Based on the detailed analysis of the space location characteristics of Location Dependent Data （LDD） and several typical replacement strategies of location dependent cache, the authors proposed a prioritized approach cache replacement based on the lowest access cost （ PLAC）, the PLAC took some important factors into account such as access probabilities, update rates, data distance, valid scope. To ensure the maximum utilization of limited cache, the PLAC cache replacement strategy decided which data would be replaced according to the value of the lowest cost function. The contrast experiments show that the PLAC increases cache hit rate and shortens query average response time more effectively than other location dependent cache replacement strategies.