PRIVACY-PRESERVING KEYWORD SEARCH WITH SYMMETRIC-KEY VERIFICATION OVER DYNAMIC CLOUD DATA

Abstract

Cloud computing has become a fundamental platform for data storage and management due to its scalability, flexibility, and cost-effectiveness. However, outsourcing sensitive information to cloud servers introduces significant security and privacy concerns, particularly when users need to perform keyword searches over encrypted data. Traditional searchable encryption techniques enable data retrieval without revealing plaintext information, but they often face challenges related to search verification, dynamic data updates, and protection against malicious or untrusted cloud servers. Ensuring that search results are both accurate and complete while preserving data confidentiality remains a critical requirement in modern cloud storage systems. This paper proposes a privacy-preserving keyword search framework with symmetric-key verification over dynamic cloud data. The framework integrates symmetric-key cryptographic mechanisms, searchable encryption techniques, and verifiable search protocols to enable secure keyword retrieval from encrypted cloud repositories. Data owners encrypt files before outsourcing them to the cloud and generate secure searchable indexes using symmetric-key operations. Authorized users can perform keyword searches without exposing sensitive information to the cloud server. To guarantee result correctness and completeness, a verification mechanism is incorporated that allows users to validate returned search results and detect any omission, modification, or forgery performed by an untrusted cloud provider. The proposed framework further supports dynamic cloud environments where encrypted files can be inserted, modified, or deleted without requiring complete index reconstruction. Efficient update procedures maintain search accuracy while minimizing computational and communication overhead. Security analysis demonstrates that the framework preserves data confidentiality, search privacy, keyword unlinkability, and verification integrity against various attack models. Experimental evaluation indicates that the proposed approach achieves efficient search performance, low storage overhead, and strong verification capabilities compared with conventional searchable encryption schemes. The framework provides a practical and scalable solution for secure cloud data management in applications requiring privacy-preserving data retrieval, including healthcare, finance, government services, and enterprise cloud storage systems.