EnDAS: Efficient Encrypted Data Search as a Mobile Cloud Service

Abstract

Document storage in the cloud infrastructure is rapidly gaining popularity throughout the world. However, it poses risks to consumers unless the data is encrypted for security. Encrypted data should be effectively searchable and retrievable without any privacy leaks, particularly for the mobile client. Although recent research has solved many security issues, the architecture cannot be applied on mobile devices directly under the mobile cloud environment. This is due to the challenges imposed by wireless networks, such as latency sensitivity, poor connectivity, and low transmission rates. This leads to a long search time and extra network traffic costs when using traditional search schemes. This study addresses these issues by proposing an efficient Encrypted DAta Search (EnDAS) scheme as a mobile cloud service. This innovative scheme uses a lightweight trapdoor (encrypted keyword) compression method, which optimizes the data communication process by reducing the trapdoor's size for traffic efficiency. In this study, we also propose two optimization methods for document search, called the Trapdoor Mapping Table (TMT) module and Ranked Serial Binary Search (RSBS) algorithm, to speed the search time. Results show that EnDAS reduces search time by 34% to 47% as well as network traffic by 17% to 41%.

Searchable encryption: A survey on privacy‐preserving search schemes on encrypted outsourced data

“…In some cases, the query may be generated by the DO on behalf of the end users. 61 The use of precomputed Trapdoor Mapping Table (TMT) to store the trapdoor corresponding to the English words is proposed by Ma et al, 62 as it significantly reduces the overhead in trapdoor generation by reducing the communication overhead between the data owner and the end user.…”

“…In some of the schemes, the index is converted into a tree structure for efficient retrieval. 16,27,29,35,36,38,[44][45][46]62,68,70,80,93,107 Table 9 lists the classification of the existing SE schemes based on the type of index.…”

Document clustering for efficient and secure information retrieval from cloud

“…This leads to increased communication overheads as irrelevant documents are downloaded by the end‐users. Index‐based searching in employed by various researchers either using forward index, inverted index, and tree‐based index …”

“…Index-based searching in employed by various researchers either using forward index, ‡ inverted index, § and tree-based index. 16,41,[64][65][66][67] Ranked Retrieval -To provide ranked results based on relevance of the search query, the introductory work was proposed by Wang et al 7 based on Order-preserving encryption. 68 This scheme was able to rank the results but was applicable to single keyword only.…”

“…These techniques are not preferred as they do not represent the end‐users search behavior as the end‐users prefer to use multiple keywords (ie, 2 to 5 keywords) in the search query . Multi‐keyword search techniques exist, but these schemes require comparisons proportional to the number of documents in collection or a universal dictionary structure to be shared with the end‐users. This increased number of comparisons leads to increased latency with which results are provided.…”