中文摘要：

针对广域网环境中现有缓存一致性机制陈旧率高，扩展性差的缺陷，提出了一种自适应的缓存一致性机制，该机制通过对数据改变频率的量化分析，从多层次的一致性策略中智能化的为数据选择一种最合适的策略。通过仿真实验表明自适应一致性机制灵活性好，陈旧率低，副本更新频率优化。

英文摘要：

Focus on the limitation of high staleness rate and low expansibility in current cache consistency mechanism of current WAN, proposed a new adaptive consistency mechanism for cache on the base of currently universal consistency mechanisms. By the analysis of data change frequency, the new mechanism chooses the best strategy automatically from the muti -level consistency mechanism strategies. The imitate experimental results showed that the adaptive cache consistency mechanism implements easier, keeps low staleness rate and optimized copy update rate.