AK2R: An Aggregate-Key–Oriented Retrieval Paradigm for Ciphertext-Only Cloud Search

Abstract

Key-aggregate retrieval allows a cloud platform, acting for authorized users, to execute keyword queries on data encrypted using multiple public keys. Although current keyaggregate searchable encryption approaches support effective data distribution, many of them are still susceptible to keyword guessing attacks, which may cause unintended exposure of keyword ciphertexts and search trapdoors. These weaknesses seriously compromise data confidentiality in cloud-driven systems, especially within developing domains such as the Internet of Things (IoT). In this work, we introduce a secure and optimized key-aggregate keyword retrieval mechanism for encrypted cloud data. The presented approach enables a data owner to flexibly distribute encrypted files to multiple recipients through a single lightweight aggregate key, while permitting legitimate users to delegate keyword search operations to the cloud server without disclosing confidential keyword details. To mitigate privacy risks, the scheme is constructed to resist keyword guessing attacks and to preserve the secrecy of both keyword ciphertexts and trapdoors. We rigorously specify the security framework of the proposed method and demonstrate its robustness under the indistinguishable selective-file chosen keyword attack (IND-SF-CKA) model. Computational evaluation and simulation-based experiments indicate that the proposed solution delivers strong security assurances with minimal computational and communication costs, rendering it suitable for cloud-enabled IoT environments and large-scale encrypted data sharing scenarios.