基于密文策略属性基加密系统访问机制的缓存替换策略

摘要

tn order to improve the performance of cache for encrypted data based on Ciphertext Policy Attribute Based Encryption (CP-ABE),an effective replacement algorithm named Minimum Attribute Value (MAV) algorithm was proposed.Combining the access mechanism of ciphertext in CP-ABE and counting the number of high frequency attribute values,the attribute similarity was calculated by using cosine similarity method and the table of high frequency attribute values;meanwhile,the attribute value of each cache file was calculated according to the attribute similarity and size of the encrypted file,then the file with the minimum attribute valuve was replaced.The experimental results prove that the MAV algorithm has better performance in increasing byte hit rate and file request hit rate than the algorithms of Least-Recently-Used (LRU),Least-Frequently-Used (LFU) and Size for encrypted data based on CP-ABE.%为提高基于密文策略属性基加密(CP-ABE)系统的数据缓存性能,针对CP-ABE加密的数据,提出一种有效的缓存替换算法——最小属性价值(MAV)算法.该算法结合CP-ABE加密文件的访问策略并统计高频属性值的个数,利用余弦相似度方法和高频属性值统计表来计算属性相似度;同时结合属性相似度和文件大小计算缓存文件的属性值价值,并替换属性值价值最小的文件.在与最近最少使用(LRU)、最不经常使用(LFU)、Size缓存替换算法的对比实验中,针对CP-ABE加密后的数据,MAV算法在提高加密文件请求命中率和字节命中率方面具有更好的性能.