Practical Single-server (Stateful) Private Information Retrieval
Ling Ren
Abstract:
In this talk, I will present our recent results on practical single-server private information retrieval (PIR) and stateful PIR. First, I will briefly cover OnionPIR (version 2), a single-server PIR protocol that significantly improves communication and server computation through a combination of FHE composition, protocol design, and engineering.
PIR schemes in the standard model (OnionPIR included) are faced with a linear server computation lower bound. I will then present two stateful PIR protocols where the client stores hints to enable sublinear server computation. The first one is lightweight and (after an offline preprocessing step) completes a query within milliseconds. The second protocol achieves the optimal trade-off between client storage and server computation.
Bio:
Ling is an assistant professor in Siebel School of Computing and Data Science at the University of Illinois at Urbana-Champaign. Prior to joining the University of Illinois, he received his PhD from MIT and worked as a postdoctoral researcher at VMware Research Group.
He works on applied cryptography and secure distributed algorithms. He is generally interested in designing algorithms that are both practically efficient and provably secure. As a result, his research has been as theoretical as proving asymptotic bounds, and as practical as taping out a processor.