The proposed schemes are secure only in a bounded model: the systems admit $L$ secret keys, at most, for a bound $L$ that must be fixed in the setup of the systems. The efficiency of the cryptosystems, later, depends on this bound $L$. Although this is an important drawback that can limit the applicability of the proposed schemes in some real-life applications, it turns out that the bounded security of our key-policy attribute-based encryption scheme (in particular, with $L=1$) is enough to implement the generic transformation of Parno, Raykova and Vaikuntanathan at TCC'2012. As a direct result, we obtain a protocol for the verifiable delegation of computation of boolean functions, which does not employ pairings or lattices, and whose adaptive security relies on the Decisional Diffie-Hellman Assumption.
Category / Keywords: cryptographic protocols / attribute-based cryptography, Discrete Logarithm setting, verifiable computation Date: received 6 Mar 2015 Contact author: jherranz at ma4 upc edu Available format(s): PDF | BibTeX Citation Version: 20150306:133422 (All versions of this report) Short URL: ia.cr/2015/213 Discussion forum: Show discussion | Start new discussion