TL;DR
This paper introduces SRAP-PRET, a novel cryptographic scheme for the Internet of Medical Things (IoMT) that offers robust security against both external data leaks and internal algorithmic attacks. It achieves this by combining autonomous path delegation, secure data deduplication, and a cryptographic reverse firewall, ensuring patient data privacy and system integrity in resource-constrained medical environments.
What Problem Does This Paper Address?
The Internet of Medical Things (IoMT) promises ubiquitous healthcare services, but it also introduces significant security challenges. Sensitive medical records are collected and shared across a network of diverse devices, making them vulnerable to data exposure from cloud servers or open networks. A more insidious threat comes from internal adversaries who can perform "algorithm substitution attacks" (ASAs), subtly altering cryptographic algorithms to steal data without detection. Furthermore, existing systems struggle with flexible delegation of data access rights, especially when data owners are unavailable, and efficient storage of increasingly large medical datasets. This paper tackles these issues by proposing a scheme that is resistant to subversion, supports fine-grained delegation, and enables secure data deduplication.
Methods
The core of the proposed solution is the SRAP-PRET (Subversion-Resistant Autonomous Path Proxy Re-Encryption with Equality Test) scheme. Key techniques include:
- Autonomous Path Delegation: Data owners can pre-define a specific, ordered path for data access delegation. This prevents unauthorized access and maintains control even in multi-hop scenarios, mitigating the trust degradation often seen in such systems.
- Cryptographic Reverse Firewall (CRF): A CRF zone is deployed on both the Key Generator (PKG) and client sides. This acts as a protective layer that sanitizes keys and messages, rerandomizing them to prevent hidden backdoors and information leakage, even if the underlying devices are compromised.
- Proxy Re-Encryption (PRE): This allows a semi-trusted proxy to transform ciphertexts from one recipient's key to another's without decrypting the data, facilitating secure data sharing.
- Equality Test (PKEET): This enables the cloud server to efficiently check if two ciphertexts correspond to the same plaintext, crucial for secure data deduplication.
- Secure Deduplication: By combining PRE and PKEET, the system can identify and eliminate redundant data, saving storage space.
Datasets
Not discussed in the paper. The paper focuses on cryptographic primitives and theoretical security proofs, with performance evaluated through simulations and comparisons with existing schemes.
Key Insights and Ideas
- CRFs as a General Defense: The paper highlights CRFs not just as a specific defense but as a versatile framework that can be integrated into various cryptographic protocols to resist internal subversion and exfiltration.
- Balancing Security and Efficiency: The authors demonstrate that it's possible to achieve strong security guarantees (subversion resistance, CPA security) without significantly sacrificing computational and communication efficiency, which is critical for resource-constrained IoMT devices.
- Autonomous Paths for Trust Management: The concept of pre-defined, autonomous delegation paths is a key innovation for managing trust in complex, multi-party systems like IoMT, where direct trust might not extend across many hops.
- Addressing the "Weakening Trust" Problem: By allowing delegators to explicitly define the delegation path, the scheme directly addresses the issue of diminishing trust as data passes through multiple intermediaries.
Key Results
- Security Guarantees: SRAP-PRET is formally proven to be resistant to Algorithm Substitution Attacks (ASAs) and achieves IND-CPA security (Chosen Plaintext Attack security) under the cryptographic reverse firewall model.
- Performance: Experimental simulations show that SRAP-PRET offers comparable or even better performance (computational and communication overhead) than existing schemes, especially in terms of trapdoor generation and overall efficiency on resource-constrained devices like Raspberry Pi. For instance, decryption on a Raspberry Pi completes within about one second.
- Storage Efficiency: The scheme achieves secure data deduplication, reducing storage costs.
Limitations and Future Work
- Complexity of CRF Deployment: While CRFs offer strong security, their deployment and management might add complexity to the system architecture.
- Potential for Branch Delegation: The authors suggest future work could explore incorporating "branch link construction technology" to further optimize decryption efficiency and handle more complex delegation scenarios beyond a linear path.
- Malleable CRF Integration: They propose exploring the integration of malleable CRFs with other advanced cryptographic techniques to create a more general framework for subversion-resistant proxy re-encryption.
Summary updated