SGX-based Big-Data Analytics Frameworks
Introduction
- Set of security-focused CPU instructions
- Protects application code and data
- Uses enclaves for secure memory execution
- Prevents disclosure/modification of sensitive data
PROBIDA Framework
Core Design
- Built on Harp-DAAL
- Supports iterative, MPI-based cluster computing
- Hybrid model: selective enclave usage
- Keeps management components outside, workers inside enclaves
Key Features
- Secure enclave creation and management
- Cross-enclave attestation
- Secure inter-enclave connections
- Automated deployment optimization
- Task graph translation tools
Capabilities
- Scalable big-data analytics
- Protected machine learning operations
- Strategic enclave deployment
- Resource-aware optimization
Publications
Conferences/Workshops
C. Widanage, W. Liu, J. Li, H. Chen, X. Wang, H. Tang, J. Fox, HySec-Flow: Privacy-Preserving Genomic Computing with SGX-based Big-Data Analytics Framework, in the Proceedings of IEEE 14th International Conference on Cloud Computing (CLOUD), September 5-11, 2021. Link