Best Paper Award

  Public-key encryption (PKE) enables private communication between a sender and a receiver without sharing secret information. PKE is fundamental for information technology. Attribute-based encryption (ABE), a type of PKE, provides a variety of access controls based on attributes and policies. Key-policy ABE embeds attributes in ciphertexts, and the policy is associated with the decryption key. In ciphertext-policy ABE, each ciphertext includes the policy, and decryption is allowed with keys having the attributes satisfying the policy. For example, in key-policy ABE, content is encrypted with attributes such as “genre = music, day = 20190501,” and allows decryption only with keys associated with a policy such as “(genre = sport) OR (genre = music).”
  In this work, the authors propose an efficient and generic framework for dealing with numerical attributes as “ranges.” By incorporating numerical ranges, a policy can specify a relation such as “day ∈ [20190401, 20200331].” Although a conventional ABE can describe such a policy using many disjunctions, the complexity increases by a linear order of the range size. The authors use the segment tree method for reducing the complexity overhead to a logarithmic order, and observe that it suffices to construct ABE for set intersection. Then, they show a generic construction of such ABEs from conventional ABEs. While several existing works also achieve sub-linear complexity overhead, they need restrictions on policy classes. This work provides ABE allowing any Boolean formulae as policies, and thus simultaneously achieves high expressiveness and efficiency.
  As described above, this paper provides ABE with both high functionality and efficiency with a generic framework. The results contribute to the development of PKE with advanced functionalities. Hence, this paper deserves the IEICE Best Paper Award.