Direct chosen ciphertext security from identity-based techniques

TLDR

The work builds on efficient identity‑based encryption schemes by Boneh–Boyen and Waters, aiming to improve CCA2‑secure cryptosystems that previously used IBE as a black box. The paper proposes a new encryption technique that achieves standard‑model CCA2 security. It directly leverages the Boneh–Boyen and Waters IBE primitives, requiring only the IBE scheme itself, and is instantiated as a full encryption system and a KEM that support hierarchical IBE and public ciphertext validity. The resulting systems are CCA2‑secure in the standard model, offer shorter ciphertexts than existing methods, match the efficiency of Boneh–Katz, and enable hierarchical IBE and non‑interactive threshold decryption.

Abstract

We describe a new encryption technique that is secure in the standard model against chosen ciphertext attacks. We base our method on two very efficient Identity-Based Encryption (IBE) schemes without random oracles due to Boneh and Boyen, and Waters.Unlike previous CCA2-secure cryptosystems that use IBE as a black box, our approach is very simple and compact. It makes direct use of the underlying IBE structure, and requires no cryptographic primitive other than the IBE scheme itself. This conveys several advantages. We achieve shorter ciphertext size than the best known instantiations of the other methods, and our technique is as efficient as the Boneh and Katz method (and more so than that of Canetti, Halevi, and Katz). Further, our method operates nicely on hierarchical IBE, and since it allows the validity of ciphertexts to be checked publicly, it can be used to construct systems with non-interactive threshold decryption.In this paper we describe two main constructions: a full encryption system based on the Waters adaptive-ID secure IBE, and a KEM based on the Boneh-Boyen selective-ID secure IBE. Both systems are shown CCA2-secure in the standard model, the latter with a tight reduction. We discuss several uses and extensions of our approach, and draw comparisons with other schemes that are provably secure in the standard model.