移动环境下支持实时事务处理的数据预取  

Data Prefetching to Support Real-time Transactions Processing in Mobile Environments

作　　者：李国徽[1] 杨兵[1] 陈辉[1] 杜建强[2] 

机构地区：[1]华中科技大学计算机学院,武汉430074 [2]江西中医学院计算机学院,南昌330006

出　　处：《计算机学报》2008年第10期1841-1847,共7页Chinese Journal of Computers

基　　金：国家自然科学基金(60873030);国家“八六三”高技术研究发展计划项目基金(2007AA01Z300)资助~~

摘　　要：随着移动通信技术的迅速发展,人们提出了新的应用要求:在移动环境下处理实时事务.而移动通信带宽有限性引起较大的数据访问延迟,有时甚至由于网络传输的断接使得事务得不到所需要的数据,数据预取能够很好地解决这个问题.已有的移动环境下数据预取没有考虑到数据的流行性和事务的时间特性.该文分析影响实时事务数据预取的因素,首先考虑数据易变性、活跃性等因素,获得高价值预取数据集合;然后考虑访问预取数据的事务优先级、数据流行性等因素,构造预取数据的选择函数,通过该函数在前面选取的集合中筛选出对满足实时事务截止期更有价值的数据对象进行预取.实验表明,该数据预取策略能降低移动实时事务满足截止期的比率,更好地支持移动实时事务处理.With the rapid advances in mobile computing technology, there is an increasing demand for processing real-time transactions in mobile environments. Due to the intrinsic limitations of a mobile environment, there can be long time latency for a data access. Sometimes the data can even be unavailable due to the network disconnection. Data prefetching is an approach to this problem. However, most of the previous studies in data prefetching don＇t consider the time- liness for data objects and deadlines of transactions. Therefore they are inefficient for supporting mobile real-time transaction processing. This paper proposes a data prefetching strategy to support real-time transaction processing in a mobile data broadcast environment. Each data object is assigned a prefetching value determined by its changeability, access frequency. For the data objects with bigger prefetehing values, by further considering the priorities of the transactions which access the data objects and the validity intervals for these data objects, the authors deter mine whether these data objects are actually prefetched. Extensive simulation experiments are conducted and show that when using the proposed data prefetching strategy, more mobile real- time transactions can meet their deadlines.

关 键 词：数据预取 移动实时事务 缓存失效 预取策略 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]