Search

Traditional blockchains cannot achieve the same transaction throughput as Web2, so their use cases are limited. Therefore, state sharding has been proposed to improve transaction throughput by dividing the blockchain network and managing states and transactions in parallel. However, Nightshade in the NEAR Protocol, a type of state sharding, provides a rollback protocol to cancel the generation of blocks containing inconsistent transaction results because processing cross-shard transactions (CSTXs) in a 2-phase commit may cause state inconsistency. We present a new attack that interferes with the generation of new blocks by repeatedly executing CSTXs that certainly causes state inconsistency, causing continuous rollback. We also propose a block-generation protocol called Synchro to incorporate all the state changes of each CSTX into the same block by coordinating the block prior to approving transactions in each shard. Synchro eliminates the occurrence of the state inconsistency caused by the CSTXs and the necessity of the rollback protocol. We use zero-knowledge proof to make Synchro scalable in the global validation phase. Although the actual overhead of the zero-knowledge proof has not yet been evaluated, we show that Synchro could achieve the same transaction throughput as Nightshade theoretically, depending on the future innovations in zero-knowledge proof techniques.

In blockchain-based anonymous cryptocurrencies, due to their tamper-resistance and transparency characteristics, transaction data are initially required to be anonymous, with the help of various cryptographic techniques, e.g., commitment schemes and zero-knowledge proofs. Also, cryptocurrencies are different from existing anonymous messaging protocols regarding the software architecture and the underlying security model. Due to these differences, the sense of anonymity must be specifically defined for anonymous cryptocurrencies, and the anonymity in each anonymous cryptocurrency must be analyzed and evaluated based on the specific architecture model. In this paper, we first propose a specific architecture model with three software layers to anonymous cryptocurrencies. Next, we introduce definitions of fundamental privacy properties (Pfitzmann's anonymity, unlinkability, and pseudonymity) and comprehensively analyze each privacy property for each architecture layer of anonymous cryptocurrencies to establish a privacy evaluation policy for anonymous cryptocurrencies. Finally, we fairly compare the privacy of current leading anonymous cryptocurrencies (e.g., Zerocash, CryptoNote, and Mimblewimble) using the privacy evaluation policy., Comment: This work has been submitted to the IEEE for possible publication