Due to the limited scalability and reliability of traditional storage systems, more and more medical data are stored in cloud. However, the cost of cloud-assisted health systems is skyrocketing. How to improve storage utilization and save resource cost for health systems has become urgent problems. Therefore, this study proposes a secure lossless redundancy elimination scheme with semantic awareness, called SLRE, to decrease storage overhead and improve the overall system performance. Compared with existing studies, when the locality of data is almost nonexistent or poor, SLRE can eliminate redundant data well by exploiting semantic properties of similarity and entropy rather than relying on the locality. Then, fine-grained post-deduplication delta compression based on proxy similarity discovers more similar blocks and improves storage efficiency. In addition, filtered Lempel–Ziv compression based on proxy entropy filters data with low redundancy to improve the overall performance. Importantly, in order to ensure the security of medical data after deredundancy, different auditing strategies for medical data with diverse characteristics are proposed to guarantee security and improve the auditing efficiency. The experimental results show that compared with the state-of-the-art methods, the overall compression ratio of SLRE is improved, and the system throughput and auditing efficiency are also improved.
Read full abstract