Feasibility and Infeasibility of Secure Computation with Malicious PUFs

Result by Dana Dachman-Soled, Nils Fleischhacker, Jonathan Katz, Anna Lysyanskaya, Dominique Schröder

Nils talks about (in)feasibility results of maliciously generated PUFs. First remember that PUF is an acronym for physically uncloneable function and can be viewed as a singleton hardware device. It takes an input, evaluates a function on this input, and returns the output. However, it is not possible to make a clone of such a device (either physically or in software). In general we wish the function of a PUF to be random so we can view it as a random oracle. One might wonder why we care about PUFs? Results have been shown that under the assumption of a correctly constructed PUF it is possible to achieve unconditional secure MPC. Furthermore, it is possible to achieve UC secure protocols without a trusted setup, like a common reference string, when using PUFs. However, in regards to feasibility results, when one assumes that the PUF can behave maliciously, results are scarce. Nils and his co-authors show that if the maliciously constructed PUF is stateless, then it IS possible to do unconditionally UC secure MPC, but if the PUF is stateful, and thus may record queries, then it is NOT possible to do unconditionally UC secure MPC. Their proofs are based on the construction of oblivious transfer from PUFs, which is sufficient for secure computation.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s