End-to-end encryption guarantees message confidentiality but does not hide metadata such as communication patterns among senders and recipients, or their identities. Oblivious Message Retrieval (OMR) is a cryptographic protocol that enables servers to assist recipients in retrieving their messages from a database without learning the mapping between messages and recipients, thereby protecting such metadata.
This paper investigates two central questions of OMR: (1) What is the precise relationship between OMR and the better-studied primitive of Private Information Retrieval (PIR)? (2) Can OMR schemes achieve concrete efficiency comparable to state-of-the-art PIR protocols?
We show that OMR with a property we call strong detection-key-unlinkability is at least as hard as PIR, and that existing OMR constructions already satisfy this property. This PIR-to-OMR reduction has low overhead, suggesting that OMR cannot be made substantially more efficient than PIR.
We then present , which achieves to faster server runtime over the state-of-the-art under practical parameter settings. For messages of 612 bytes each, completes in only seconds with 4 MB of communication, compared to seconds and 260 KB for . These gains come with two trade-offs: an asymptotically linear digest size (albeit with small constants), and two rounds of interaction between the detector and the client.
Furthermore, crucially, uses batch PIR as a black-box component, which in our experiments accounts for -- of the server runtime. Thus, nearly matches the aforementioned lower bound concretely (for databases of to messages, each with to bytes), given the status quo of batch PIR.