A Dynamic Scheme for Secure Searches over Distributed Massive Datasets in Cloud Environment using Searchable Symmetric Encryption Technique

Özet

Cloud computing has produced a paradigm shift in large-scale data outsourcing and computing. As the cloud server itself cannot be trusted, it is essential to store the data in encrypted form, which however makes it unsuitable to perform searching, computation or analysis on the data. Searchable Symmetric Encryption (SSE) allows the user to perform keyword search over encrypted data without leaking information to the storage provider. Most of the existing SSE schemes have restrictions on the size and the number of index files, to facilitate efficient search. In this paper, we propose a dynamic SSE scheme that can operate on relatively larger, multiple index files, distributed across several nodes, without the need to explicitly merge them. The experiments have been carried out on the encrypted data stored in Amazon EMR cluster. The secure searchable inverted index, organized as a HashMap, is created instantly using Hadoop MapReduce framework during the search process, thus significantly eliminate the need to store keyword-document pairs on the server. The scheme allows dynamic update of existing index and document collection. The parallel execution of the pre-processing phase of the present research work reduces the processing time at the data owner. An implementation of our construction has been provided in this paper. Experimental results to validate the efficacy of our scheme are reported.