PIANO: Extremely Simple, Single-Server PIR with Sublinear Server Computation
Mingxun Zhou, Andrew Park, Elaine Shi, Wenting Zheng
IEEE Symposium on Security and Privacy 2024 · Day 3 · Continental Ballroom 6
This article delves into PIANO, a groundbreaking Private Information Retrieval (PIR) construction presented at IEEE S&P. The talk, led by Mingxun Zhou, a fourth-year PhD student from Carnegie Mellon University (CMU), alongside collaborators Andrew Park, Elaine Shi, and Wenting Zheng, introduces a novel approach to single-server PIR that achieves sublinear computation cost per query. The core problem PIANO addresses is the pervasive privacy leakage inherent in standard information retrieval processes, such as DNS lookups, browsing history, and web searches, where a database server learns sensitive details about a user's query.
AI review
PIANO isn't just another academic paper; it's a genuine breakthrough in practical single-server Private Information Retrieval. Achieving sublinear query costs with an elegant, simple design, it makes truly private lookups viable for billions of records, effectively shifting the bottleneck from computation to network. This is the kind of foundational work that actually moves the needle for privacy-preserving systems.