DLT Interoperability and More ⛓️#17 — Generic Privacy Preserving Private Permissioned Blockchains⛓️
In this series, we analyze papers on blockchain and interoperability.
This edition covers a paper on a paper that proposes anonymity mechanisms for private blockchains.
➡️ Title: Generic Privacy Preserving Private Permissioned Blockchains
➡️ Authors: Frédéric A Hayek, Mirko Koscina, Pascal Lafourcade, Charles Olivier-Anclin
➡️ Paper source: https://hal.uca.fr/hal-03906880/document
To read this paper, basic background on privacy is useful. Quoting from the paper:
“A good example of pseudonymity is Bitcoin , where users are identified by their public keys, and it is generally infeasible to trace the public key to a real-world identity. Note that in bitcoin a user can create as many pseudonyms as they wish. Anonymity, on another hand, is when it is infeasible to link anything. One such example is Monero’s usage of ring signatures  in order to anonymize the sender’s identity that is blended with a set of other identities”.
- The authors propose a privacy-enhanced adaptation of Hyperledger Fabric powered by a BFT consensus algorithm.
💪 Strong points:
- Good work on the background on signature schemes, which helps understand the paper in detail. In particular, the authors use blind signatures, group signatures, and ring signatures.
- The PBFT-based algorithm inspired by Fabric is well-described.
- The paper explains in simple terms how can anonymity and pseudonymity be achieved for user transactions and ordered / endorser operations.
🤞 Suggestions for improvement:
- It would be interesting to see a formal analysis of the proposed protocols, e.g., using the UC framework.
🔥 Points of interest:
- Blind signatures and group signatures are used for anonymity. They rely on authority.
- Looks like, structurally, the proposal relies on Hyperledger Fabric with minimal changes.
🚀 What are the implications for our work?
Understanding techniques for privacy and security allow us to design more dependable blockchain interoperability middleware.