基于Markov过程预测的页面置换方法

New OS paging algorithm based on Markov process prediction

页面替换是操作系统实现虚拟存储的关键,使用的方法很大程度上影响系统效率。分析了传统页面替换方法的缺点,提出一种基于Markov随机过程的页面调度模型。通过分析页面置换序列,发现其规律并模拟该序列,目的是接近最优替换方法以降低缺页率。实现了基于该模型的页面置换方法,通过实验证明新方法能提高页面换入命中率并控制时间耗费。

The page replacing was the key mechanism for operation systems to realize the virtual memory storage. The methods adopted impact the systems efficiency greatly. After analyzing disadvantages of traditional page replacing methods, a new paging model based on Markov random procedure was proposed. By analyzing the paging sequence, the model could discover the regulation and stimulate this sequence. The purpose was to approach the optimum method and decrease the page missing rate. This paging method based on the referred model was implemented. Experiments show that this new method can increase page hitting rate and control time cost.