Fully secure hidden vector encryption under standard assumptions

Abstract

Fully secure hidden vector encryption under standard assumptions AbstractHidden Vector Encryption (HVE) is a special type of predicate encryption that can support conjunctive equality and range searches on encrypted data. All previous HVE schemes were proven to be either selectively secure or weakly attribute-hiding. In this paper, we first construct a new HVE scheme that is fully secure under standard assumptions. Our HVE scheme, which is based on bilinear maps (pairings), provides efficiency advantages in that it requires O(1)-sized private keys and O(1) pairing computations for decryption, regardless of both the number of conjunctives and the dimension of vectors. To achieve our goal, we develop a novel technique to realize a tag-based dual system encryption in prime-order groups and show how to hide vector components and compress tag values into one.

New chosen-ciphertext secure identity-based encryption with tight security reduction to the bilinear Diffie–Hellman problem

“…Because the DBDH assumption is weaker than the 172 Decision Linear assumption in groups with symmetric bilinear maps [30], we have that ≤ (q + 2) , where is the maximum 173 value among { i } q i=0 and f under the Decision Linear assumption. Therefore, Waters' tag-based IBE system does not achieve a 174 tight security reduction overall, even though each transition has a tight reduction.…”

CloudKeyBank: Privacy and Owner Authorization Enforced Key Management Framework

“…As is defined in Definition 3.1, HVE [6][10] [11] is a type of predicate encryption where two vectors x and w over attributes are chosen, x is associated with a ciphertext, and w is associated with a token. The ciphertext matches the token if and only if the two vectors meet with Equation (1).…”

“…In this section, based on known schemes HVE[10] [11] and PRE [29], we present a concrete SC-PRE scheme which provides the privacy and authorization enforced encryption support for our CloudKeyBank framework by distributing its eight polynomial algorithms into the following two protocols: DepositKey protocol described in Section IV-B and With-drawKey protocol describe in Section IV-C.…”

“…-If both of the two above are false, then it outputs ⊥ to the ReEnc algorithm. Search algorithm is the first level of access control enforcement by the CloudKeyBank Provider because of the following truth: to protect search privacy on identity attributes against the honest-but-curious CloudKeyBank provider, Search algorithm is implemented based on the fact described in HVE [11] [10] where two vectors x l ∈ W l and w i ∈ W * over identity attributes are chosen in SC-PRE. Therefore if evaluating QuT o, wi against ciphertext CT l ∈ EDB makes the P wi ( x l ) = 1, Search algorithm outputs ek i ||f lag, otherwise outputs ⊥.…”

“…It still results in the huge query cost because each of the ciphertexts could be a valid ciphertext that matches the token. Park proposed a low cost HVE [10] [11] scheme which is also based on pairing group but only requires a O(1) token size and O(1) pairing computation. Therefore following predicate privacy in [9] and scheme proposed in Park [10] [11], based on search attribute group x i of key tuple t i we introduce two vectors to guarantee the search privacy on conjunctive identity attributes.…”