A New Leakage Resilient Symmetric Searchable Encryption Scheme for Phrase Search

Abstract

: Symmetric searchable encryption (SSE) schemes are preferred over asymmetric ones for their lower computa-tional cost. Owing to the big data size of most of the cloud applications, SSE with keyword search often yields a large number of search results matching the search criterion, but only a small portion of them is of actual interest. This results in unnecessary increase of network traffic. A customized search against a phrase instead of keywords can yield more specific and relevant search results and can reduce the network traffic. This motivates the idea of phrase search in SSE. Most of the existing symmetric key searchable encryption schemes either do not support phrase search or have unwanted leakage associated with them. In this paper, we propose a symmetric key searchable encryption scheme for phrase search that minimizes the leakage of information from search pattern and access pattern . We propose a probabilistic trapdoor generation algorithm for phrase search and thereby prevent the leakage due to search pattern . In earlier SSE based schemes, an honest-but-curious server could always learn about the position of the sentences and keywords in the encrypted text after the search operation is performed. This is referred to as the leakage from access pattern . This may turn out to be a significant security concern owing to the prior knowledge of positions of certain sentences and keywords in certain documents. In this paper, we provide the access pattern secure encryption scheme such that, an honest-but-curious cloud server could not learn anything about the position of the phrase in the sentence even after the search. We implement a prototype of our scheme and validate it against commercial data and provide security and performance analysis to demonstrate its practicality.