Ethereum co-founder Vitalik Buterin has co-authored a research paper which explains how sharding can be implemented on the Ethereum blockchain without compromising its security. Published on September 24th, the paper includes contributions from University College London Phd students Mustafa Al-Bassam and Alberto Sonnino.
Maximizing SPV Security
In the paper titled “Fraud Proofs: Maximizing Light Client Security and Scaling Blockchains with Dishonest Majorities”, the authors note that the Ethereum network can be more secure if its codebase implements (or incorporates) a new type of mathematical proof before the sharding upgrade.
SPVs are nodes on a blockchain that only download a “small portion” of network data and usually validate blocks by assuming (or trusting) that the blockchain’s consensus algorithm “only contains valid blocks.”
Conforming to "Protocol Rules", "Probabilistic Sampling"
Downloading only a portion of blockchain data and validating blocks based on only a partial history of transaction logs makes a strong assumption that the “majority of block producers are honest.”
Al-Bassam, Connino, and Buterin propose that by sending "fraud proofs" to SPVs, the “honest-majority assumption” can be eliminated. They explain that fraud proofs sent by full node operators will allow network participants to make “much weaker assumptions about a minimum number of honest nodes that rebroadcast data.”
The researchers claim that these fraud proofs will accurately determine and show whether a block does not conform with the network’s “protocol rules.” Additionally, “probabilistic sampling techniques” will be used to determine whether block data “is available to be downloaded.”
Constantinope Upgrade, zk-SNARKS
The authors assert that effective on-chain blockchain scalability solutions such as sharding or increasing block sizes may be implemented by using fraud and data availability proofs detailed in their research paper.
As covered on CryptoGlobe, Buterin recently noted that Zcash’s (ZEC) zk-SNARK technology could be used to help Ethereum scale to process 500 transactions per second (TPS). Currently, its network is only able to handle (an average of) 15 TPS.
Buterin’s exploration of new ideas comes at a time when the Ethereum platform is preparing for a major upgrade referred to as Constantinople. As CryptoGlobe reported, the Constantinople (hard fork) upgrade will involve incorporating five different Ethereum Improvement Proposals (EIPs) - which aim to improve the Ethereum blockchain’s network efficiency and make modifications to its fee structure (for miners).