Research On Synonym Search And Leakage-resilient In SSE

In an era of mass data, more and more users start to utilize cloud outsourcing services of cheap and powerful computing ability. While security problem has become a main obstacle to its further moveforward. The fact is that cloud server provider is”honest-but-curious”. A naive method is that these data have to be encrypted prior to outsource to the remote semi-trusted servers to combating unsolicited accesses. But unfortunately, this way restricts the users' effective search ability on encrypted data.Therefore, this point that how to avoid accessing the outsourced data by unauthorized users while maintaining effective computing on encrypted data has been and advanced research topics in the field of cloud outsourcing.The searchable encryption technique based on secure index, which has been a key and lead solution, solves this issue. It mainly consists of symmetric searchable encryption(SSE), public-key searchable encryption(PKSE) and private information retrieval(PIR). And they, respectively, work out the application problems under the different scenarios. In this thesis, we only focus on SSE technique in the context of cloud computing. The dissertation mainly researchs on problems of data privacy leakage and synonym search on encrypted data in untrusted cloud setting.In order to solve the issue of information leakage in the existing SSE, we present a searchable encryption scheme of resisting leakage. In this thesis, we firstly define the new concepts of access pattern and information leakage. Compared with the previous single keyword search works, not only does our solution avoid the leakage of size pattern, but also it leaks access pattern and search pattern in probability. In addition, this scheme doesn't increase the computing and storage overheads of clients and communication cost between clients and servers. We prove that this scheme is Non-adaptive secure by security analysis. But it is unfortunate that this scheme greatly improves the storage and computational overheads of servers.So far, there exists no such solution focusing on the synonym search in the setting of complexity searchable encryption. In consideration of finiteness of people's memory and people's frequent changing, the existing schemes seem to be difficult to approach this case. To this case, we seriously study this issue. Soon afterwards, we for the first time put forward the issue of the synonym keyword search over remotely encrypted data in the context of cloud computing and design a formal Practical Synonym Symmetric Searchable Encryption Solution(PSSSE) in this thesis. Our proposed technique have a number of crucial advantages. In this scheme, we introduce the concepts of synonym function and synonym sets to obtain better scalability. Through rigorous security analysis, we show that our proposed solution is correct and privacy-preserving(Nonadaptive security) without the loss of function of the scheme. Besides, the algorithms that we present are effective, they almost increase no space and computing overheads of users. The overhead of communication still remains O(1). The time complexity of searching on encryption data is O(p)(p denotes the maximum number of synonym set of keyword). This scheme makes up the defects of fuzzy search and enhances system's usability by responding all documents containing the keyword and corresponding synonyms.