摘要
给出了一种树型WEB缓存结构,并描述了它的基本设计原理。在本结构中,缓存被当作一系列低通滤波器,滤波器的低截止频率等于特征时间的倒数。当文件的访问频率低于该截止频率时,将被替换;相反,将被缓存。仿真研究表明,所提出的结构和传统的非协同式结构相比,节省了缓存资源,降低了缓存的失效比。该结构已被用于作者开发的基于WEB的远程虚拟仪器实验室,明显提高了用户的访问速度。
This paper presents the tree-like WEB cache architecture, and describes its fundamental design principle. In this system, cache can be considered as a series of low pass filters, and cutoff frequency of filters is the reverse of characteristic time. When the access frequency is lower than this cutoff frequency, the document will be replaced by new document, otherwise, will be cached. The simulation results show that the presented architecture compared with the traditional architecture, can save cache resouce, and reduce the cache miss ratio. This architecture is already used in Remote Virtual Laboratory based on WEB, and the user-visiting rate is improved in evidence.
出处
《系统仿真学报》
CAS
CSCD
2002年第11期1559-1562,共4页
Journal of System Simulation