SFGKA: A Scalable and Fine-Grained Group Key Management Scheme for Cloud Computing

The dynamic nature of group collaboration in cloud computing environments demands scalable, privacy-preserving, and fine-grained key management solutions. However, existing group key agreement (GKA) protocols struggle to simultaneously support dynamic membership, attribute-based access control, and anonymity, resulting in limited applicability to modern cloud systems. To overcome these challenges, we propose DAGKA, a dynamic attribute-based group key agreement framework designed for scalable and privacy-aware group communication. DAGKA integrates attribute-based encryption with a two-level access tree structure to enforce fine-grained policies while concealing user identities. A lightweight key update mechanism, powered by polynomial aggregation, allows members to join or leave the group without triggering full rekeying. In addition, a semi-trusted server assists in access tree computation, offloading cryptographic costs from resource-constrained devices. Security analysis proves that DAGKA achieves IND-CCA security and ensures both forward and backward secrecy. Experimental results demonstrate that our scheme is superior to its counterparts, achieving constant-time computation during dynamic updates in realistic settings.