An Efficient Privacy-Preserving Ranked Multi-Keyword Retrieval for Multiple Data Owners in Outsourced Cloud

Aim:

          The aim of this study is to develop an efficient and secure privacy-preserving ranked multi-keyword retrieval scheme for encrypted cloud storage that ensures data confidentiality, resists keyword guessing attacks, and improves the accuracy and efficiency of ranked search results.

Abstract:

          With the increasing adoption of cloud storage by both individuals and organizations, data providers frequently offload their data to the cloud to alleviate memory constraints and enable rapid data retrieval, which has become a growing trend. Ensuring the confidentiality of this data has led to the development of various encrypted cloud storage solutions for ranked multi-keyword searches. However, many existing approaches are vulnerable to keyword guessing attacks, and the ranked top-K search results retrieved from encrypted cloud data are often inaccurate.

          To address these issues, we propose a new and efficient privacy-preserving ranked multi-keyword retrieval scheme (PRMKR). In PRMKR, data providers can securely transfer their encrypted data and corresponding inverted indexes to the cloud. Registered users can then perform accurate searches without revealing their trapdoor information to the cloud server. Our approach introduces an encrypted searchable plugin server and lower-dimensional inverted index vectors, which enhance both data confidentiality and search efficiency. Security analysis demonstrates that PRMKR successfully resists keyword guessing attacks, while experimental results confirm its strong computational and communication efficiency.

Existing System:

        In the existing cloud storage systems, data providers often encrypt their data before uploading it to the cloud to ensure confidentiality. Many retrieval schemes have been developed to enable ranked multi-keyword searches over encrypted data. These schemes allow users to perform searches without decrypting the data, providing a layer of privacy. However, most of the current systems suffer from vulnerabilities such as keyword guessing attacks, where an adversary could infer the search queries. Additionally, the ranked top-K search results provided by these systems are often inaccurate, leading to inefficient retrieval and compromised user experience. As a result, existing systems struggle to balance security and search accuracy while maintaining performance efficiency.

DisAdvantages:

Initial Setup Complexity: The system requires a sophisticated encryption and indexing setup, which could be complex for non-expert users or small organizations to implement.

Computational Overhead: Despite improvements, encryption, decryption, and privacy-preserving mechanisms may still introduce computational overhead compared to unencrypted data retrieval systems.

Dependence on Cloud Provider: Data providers must trust the cloud provider to maintain system integrity and avoid potential attacks on the encrypted data infrastructure.

Increased Storage Requirements: Storing encrypted data and its corresponding index vectors may require more storage space than unencrypted systems, especially for large datasets.

Proposing System:

       The proposed system introduces a novel and efficient privacy-preserving ranked multi-keyword retrieval scheme (PRMKR) for encrypted cloud storage. In this system, data providers securely transfer their encrypted data and optimized inverted index vectors to the cloud server. The system enhances data confidentiality by using an encryption searchable plugin server that safeguards the data and ensures secure search operations.

       To improve retrieval efficiency for frequently searched queries, the system integrates encrypted caching at the plugin server, allowing faster document ID retrieval without compromising security. This caching mechanism stores frequently accessed document IDs securely, reducing retrieval time for repeated searches.

       Registered users can perform accurate multi-keyword searches over the encrypted data without exposing their search terms (trapdoor information) to the cloud server. The system efficiently resists keyword guessing attacks through its rigorous security mechanisms, ensuring that the search queries remain private. Moreover, by leveraging lower-dimensional inverted index vectors and encrypted caching, the proposed system improves search efficiency and provides accurate ranked top-K search results.

        In summary, PRMKR offers a balance between data security, search accuracy, and computational efficiency, addressing the limitations of existing retrieval systems in cloud storage environments.

Advantages:

Enhanced Security: The proposed system resists keyword guessing attacks, ensuring that both the data and search queries remain confidential.

Accurate Retrieval: By utilizing optimized inverted index vectors, the system provides highly accurate ranked top-K search results, improving user experience and data retrieval precision.

Privacy-Preserving: Users can search the encrypted data without revealing their trapdoor information to the cloud server, maintaining search privacy.

Efficient Search Performance: The introduction of lower-dimensional inverted index vectors enhances the computational and communication efficiency of the search process.

Scalability: The system is designed to handle large datasets in the cloud, making it scalable for widespread adoption by organizations and individuals.