Orochi to Present RisePIR at CANS 2026
We are pleased to announce that
“Incremental Keyword Private Information Retrieval from d-ary Segmented Cuckoo Filters,” a research paper by Orochi Network researcher Bao Ninh, has been accepted for presentation at the 25th International Conference on Cryptology and Network Security (
CANS 2026).
The International Conference on Cryptology and Network Security (CANS) is a well-established forum for research in cryptography, cybersecurity, and network security. CANS brings together researchers from both cryptography and systems/network security, with an emphasis on work that bridges theory and practice.
Bao Ninh's paper is one of 41 selected from 141 submissions at CANS 2026. He will present the work in Wollongong, Australia, this November, with publication in Springer’s Lecture Notes in Computer Science proceedings.
About the Research
Excerpts from the abstract of “Incremental Keyword Private Information Retrieval from d-ary Segmented Cuckoo Filters”:
We introduce RisePIR, the first preprocessing keyword private information retrieval (PIR) scheme that absorbs insert, update, and delete on its key-value store at a cost proportional to the number of mutations alone. On a store of about one million keys, both variants patch their hint in place in 0.06–4.7 ms per mutation, five to six orders of magnitude faster than the 56–686 s full re-run of the preprocessing phase.
We deploy RisePIR-S as a privacy-preserving eth_getBalance service over the 2.05 × 10⁸ accounts of Ethereum mainnet, so that a wallet reads a balance without revealing to its RPC provider which account it holds. The running service answers a private query in 0.69 s and absorbs each block’s changes in 5.6 ms, well inside the 12 s block interval. The filter and RisePIR ship as an open-source Rust implementation, together with the benchmark harness behind every number in this paper.
The full version of the paper, which includes full proofs of the stated theorems and lemmas and additional details on parameter selection, is available on the
IACR Cryptology ePrint Archive. The eth_getBalance service is public at
risepir.org, and the implementation is open source on
GitHub.
A Closer Look at RisePIR
Private Information Retrieval (PIR) is a protocol that allows users to query data from a server without revealing which data they are interested in. PIR is very useful for systems that need privacy. Our motivation for this research comes from the observation that many use cases needing PIR for privacy involve key-value stores, and in real life, databases continuously evolve, yet no existing PIR scheme that answers keyword queries can efficiently handle this database evolution. RisePIR was born to close this gap.
Technically, RisePIR is built on two building blocks: a d-ary Segmented Cuckoo Filter and an efficient Updatable Index PIR scheme. We construct the d-ary Segmented Cuckoo Filter, a new variant of the Cuckoo Filter, to efficiently propagate edits at the key-value store level into the update operations that the underlying index PIR scheme already supports efficiently. What's special here is the elegant design of the d-ary Segmented Cuckoo Filter, which keeps this reduction's overhead small, letting us ultimately inherit the efficiency of the underlying index PIR scheme.
The result is both a theoretical and a practical contribution. On the theory side, RisePIR gives a generic construction that turns any updatable index PIR into an incremental keyword PIR (to our knowledge, the first such construction), so it isn't tied to one specific underlying scheme. We back this with a real implementation and concrete performance numbers. We've also deployed a proof-of-concept service for a real-world use case on Ethereum. You can try it yourself at risepir.org.
Continuing the Research
Bao Ninh will present RisePIR at CANS 2026 this November, sharing the work with fellow cryptography and security researchers in Wollongong, Australia. The presentation will be an opportunity to discuss the findings and hear perspectives from others working in the field.
The full paper, Rust implementation, and benchmark harness make the research available beyond the conference. Researchers and developers can examine the proofs, review the code, and reproduce the reported benchmarks.
Orochi Network continues to invest in research at the intersection of cryptography, privacy, and verifiable data. We will share further details about Bao Ninh’s presentation as the conference approaches.