UCLA

Non-interactive key exchange (NIKE) is a fundamental notion in Cryptography.

This notion was introduced by Diffie and Hellman in 1976. They proposed the

well-known 2-party NIKE protocol and left open the generic question of whether

NIKE could be realized in the multiparty setting. NIKE has since then been an

active area of research with an ultimate goal of obtaining best possible security in

the multiparty setting. Although this has evaded researchers for many decades,

advancements have been made through relaxations in multiple directions such as

restricting to 3-parties, static/semi-static model (where the adversary needs to

commit to the set of parties he wishes to be challenged upon ahead of time),

random-oracle model, allowing initial setup, etc.

This dissertation provides a solution to the open question: it provides a multiparty

NIKE protocol that is adaptively secure with no setup and in the standard

model.

The solution employs novel techniques of using indistinguishability obfuscation,

which are interesting in their own right and which seem promising in finding

wider applications in other settings. One such technique pertains overcoming the

somewhat inherent drawback of non-adaptivity of the puncturing technique introduced by Sahai and Waters [STOC'14].