Huijia Rachel Lin


Concurrent Security


Monday, Feb. 14, 2011, 4:00pm


5130 Upson Hall


On the Internet millions of applications are executed simultaneously---i.e., concurrently---unaware of each other. This concurrency brings forth new security challenges: The concurrent setting allows for a coordinated attack in which an adversary plays different roles in various protocols attempting to create rogue interactions between them.

In this talk, I will present the first general technique for achieving concurrent security without any trusted set-up while relying on standard cryptographic hardness assumptions. In essence, achieving concurrent security reduces to achieving a particular cryptographic task called a *non-malleable commitment*. Non-malleable commitments were introduced by Dolev, Dwork and Naor in 1991 and have since been extensively studied. I will also present the first constant-round construction of such a protocol based on minimal cryptographic hardness assumptions