Ciphertext policy attribute-based encryption supporting unbounded attribute space from R-LWE

Ciphertext policy attribute-based encryption (CP-ABE) is a useful cryptographic technology for guaranteeing data confidentiality but also fine-grained access control. Typically, CP-ABE can be divided into two classes: small universe with polynomial attribute space and large universe with unbounded attribute space. Since the learning with errors over rings (R-LWE) assumption has characteristics of simple algebraic structure and simple calculations, based on R-LWE, we propose a small universe CP-ABE scheme to improve the efficiency of the scheme proposed by Zhang et al. (AsiaCCS 2012). On this basis, to achieve unbounded attribute space and improve the expression of attribute, we propose a large universe CP-ABE scheme with the help of a full-rank differences function. In this scheme, all polynomials in the R-LWE can be used as values of an attribute, and these values do not need to be enumerated at the setup phase. Different trapdoors are used to generate secret keys in the key generation and the security proof. Both proposed schemes are selectively secure in the standard model under R-LWE. Comparison with other schemes demonstrates that our schemes are simpler and more efficient. R-LWE can obtain greater efficiency, and unbounded attribute space means more flexibility, so our research is suitable in practices.