ZK Fraud Proofs
Fraud proofs are retroactive disputes submitted by validators to contest the validity of a state update. They prove that fraud was committed at a certain transaction and state transition.
Traditional optimistic rollups make use of interactive fraud proofs that simulate state transitions directly on-chain. This is both inefficient, and unsafe. Inefficient because the process of submitting interactions between prover and verifier is extremely time and compute costly on Ethereum. This method of proving leads to the long 7 day withdrawal periods1. Second, the engineering overhead of making sure that the simulated execution on-chain is exactly the same as the off-chain execution environment is extremely burdensome, and can introduce subtle attack vectors2.
Layer N solves these problems by introducing a novel zero-knowledge based fraud-proving mechanism.
The process for a verifier detecting and reporting fraud on Layer N is as follows:
Verifier replays state on local machine after reading transaction level data posted on data availability layer;
Verifier detects that state root update posted to Ethereum does not match the state root that is obtained from DA replay;
Verifier calls fault proof contract on Ethereum to freeze the deposits and withdrawals on the rollup, and posts a significant stake;
The Ethereum contract now initiates a dispute period during which the verifier runs the state simulation on a RISC-V verifiable zero-knowledge virtual machine, that returns a cryptographic receipt as a proof of fraud. We've partnered with RISC Zero ↗ (opens in a new tab) to accomplish this.
The verifier then submits the receipt directly to the on-chain verifier that uses the program hash to verify the proof.
If this proof is submitted to the contract and it proves fraud, the Ethereum contract will go into state recovery mode.
Eventually, using Bonsai network developed by RISC Zero, Layer N will be able to fully transition from optimistic rollup to zero knowledge one removing the need for fraud proof period and offer privacy.