Cryptographic Reverse Firewalls for Identity-Based Encryption

The Snowden revelations show that powerful attackers can compromise user’s machines to steal users’ private information. At the same time, many of the encryption schemes that are proven to be secure in Random Oracle Model (ROM) may present undetectable vulnerabilities when implemented, and these vulnerabilities may reveal a users’ secrets, e.g., the machine hides some backdoors without the user’s awareness, and an attacker can steal the user’s private information through these backdoors. Recently, Mironov and Stephens-Davidowitz proposed cryptographic reverse firewall (CRF) to solve this problem. However, there is no CRF for identity-based encryption (IBE) has been proposed. In this paper, we propose two CRF protocols for IBE. One is a one-round encryption protocol with CRF used on the receiver, and the other is a two-round encryption protocol with CRFs deployed on both sender and receiver. We prove that these two protocols can resist the exfiltration of secret information and one is only secure against a chosen plaintext attack (CPA), the other is semantically secure against an adaptive chosen ciphertext attack (IND-ID-CCA). Moreover, we use JPBC to implement our protocols. The experimental results indicate that our protocols have some advantages in communication cost. Under certain computation cost conditions, our protocols are efficient and practical.