Skip to content
This repository has been archived by the owner on Feb 4, 2022. It is now read-only.

miguelmota/zksnarks-example

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 

Repository files navigation

zkSNARKs Example

An example of how generate zero-knowledge proofs and verify using an Ethereum smart contract.

Tutorial

  • Install circom and snarkjs:

    npm install -g circom
    npm install -g snarkjs

    NOTE: for this demo I used circom@0.0.24 and snarkjs@0.1.11

    • Circom is a language for writing arithmetic circuits that can be used in zero-knowledge proofs. Circom stands for circuit compiler.
    • Snarkjs is a JavaScript implementation of zkSNARK schemes which is used to generate proofs given circuits generated with circom.
  • Create a circuit:

    We'll create a circuit that tries to prove we are able to factor a number.

    • The circuit will have 2 private inputs and 1 output.
    • The output signal c must be the value of signal a times signal b.
    • We instantiate the circuit and assign it to the main component. The component main must always exist.

    circuit.circom

    template Multiplier() {
      signal private input a;
      signal private input b;
      signal output c;
      c <== a*b;
    }
    
    component main = Multiplier();

    Ths only thing the circuit enforces is that c equals a times b with the statement c <== a*b which defines the constraint.

  • Compile the circuit with circom:

    circom factor/circuit.circom -o circuit.json

    This will output circuit.json

  • Run setup for the circuit:

    Running the setup method will instantiate key files we'll need:

    snarkjs setup --circuit=circuit.json

    The setup will generate a proving key and a verification key, proving_key.json and verification_key.json respectively.

  • Calculate a witness:

    We'll create a file containing the input parameters.

    • Provide a file with the inputs and it will execute the circuit.
    • Snarkjs will calculate the intermediate signals and the output.
    • The calculated set of signals is the witness.

    For this example, we'll be proving we can factor 56 without revealing the two factor numbers.

    We'll prove that we know that 7 and 8 multiply to 56:

    input.json

    {
      "a": 7,
      "b": 8
    }
    • The zero knowledge proofs prove that we know a set of signals that match the constraints.
    • These signals are the witness.
    • The witness doesn't reveal any of the signals except the public inputs and the outputs.

    We can calculate the witness now:

    snarkjs calculatewitness --circuit=circuit.json --input=input.json

    This generates witness.json with all the signals.

  • Generate proof:

    We'll use the proving key and witness to generate a proof that we the inputs that factor into 56:

    snarkjs proof --witness=witness.json --provingkey=proving_key.json

    This generates proof.json and public.json which is the values of the public inputs and outputs. In this case it's just the output because our inputs were private.

    [
     "56"
    ]
  • Verify proof:

    We can verify a proof using snarkjs:

    $ snarkjs verify --proof=proof.json --verificationkey=verification_key.json --public=public.json
    OK

    This will output OK if correct or INVALID if incorrect.

  • Smart contract to verify proofs:

    snarkjs provides a method to generate a contract to verify proofs onchain:

    snarkjs generateverifier --verificationkey=verification_key.json

    This outputs verifier.sol

    The contract method we'll be using is verifyProof:

    function verifyProof(
            uint[2] a,
            uint[2] a_p,
            uint[2][2] b,
            uint[2] b_p,
            uint[2] c,
            uint[2] c_p,
            uint[2] h,
            uint[2] k,
            uint[1] input
        ) view public returns (bool r) {
        Proof memory proof;
        proof.A = Pairing.G1Point(a[0], a[1]);
        proof.A_p = Pairing.G1Point(a_p[0], a_p[1]);
        proof.B = Pairing.G2Point([b[0][0], b[0][1]], [b[1][0], b[1][1]]);
        proof.B_p = Pairing.G1Point(b_p[0], b_p[1]);
        proof.C = Pairing.G1Point(c[0], c[1]);
        proof.C_p = Pairing.G1Point(c_p[0], c_p[1]);
        proof.H = Pairing.G1Point(h[0], h[1]);
        proof.K = Pairing.G1Point(k[0], k[1]);
        uint[] memory inputValues = new uint[](input.length);
        for(uint i = 0; i < input.length; i++){
            inputValues[i] = input[i];
        }
        if (verify(inputValues, proof) == 0) {
            return true;
        } else {
            return false;
        }
    }

    As you can see there is a lot going on. Snarkjs conveniently provides a method to generate the transaction input parameters.

    snarkjs generatecall --proof=proof.json --public=public.json

    The output will look similar to this:

    ["0x225074f4c52d6abfbe39070488e9ad4a614c17e954f6dd1f95df298a0eef0a6c", "0x16c7e4b0c01a120fe48c8eb2339a7209e77796328a61e56547b73567016fd75e"],["0x250a2c4eb6c8d434b3fa3c408c1a3ab4a2db6d63c7ba917c9849083c86812832", "0x2f9c617819fec52b3417745eee59319eb56d846b5e428cb81d90a9a1f09beaf4"],[["0x19ab9c29a7e580507cc3054f36f12506d429112faa75e751ddc4c6237344ab1a", "0x099afd70d77234c35b15112b7a9352f6ec2b3030408125b66654631de21c078e"],["0x2d67175e322ab6e6a98921e29c94cdd1ab07db28d62de7e104199361e5092851", "0x0bdc9fd000cdbc79b344fb307c1ea679b2a14f1756a47cbccefe6349265fdd9d"]],["0x2f17a8a96f4f1a37dcf99fceddbb942e94a3a83547b80f4061eaa295d969dccc", "0x086cc7c64c083849eba6161be31f794b359e735e83c25a7ee9475d9943d9b8f5"],["0x26f172ffa1f027799863c43c640ccb632fc664c918138fea026127bdf07c6c0a", "0x0e0dd693317beda6b8bf1ac9024af15a35bee071cc5bdcb2784926df18ba3fb0"],["0x1c23be59028e8095a430b2e5b3768303ec5163a49e2a9b4da7f3a6b30b78a5f1", "0x0318a6a3bd9a7c36e35fc44030fda9dd594965b1cba90c942211028c3f308fb6"],["0x1f84acf96a6d7b112c097feb438dadd8bc56c59f5e466e7d17e33560df87c893", "0x257a236d53fb9a129d5e8239d592bc3173f99ab378df5f5c5a00d4e3fb6983b2"],["0x1c55a602b2237a439ccc899ae6a49e1d08ca66300283500188e5823d246a9321", "0x1c297b898c386684bc32e504317dbfdb09971227d413541f74659fcc2cd8a9a5"],["0x0000000000000000000000000000000000000000000000000000000000000038"]

    The verify function returns true if the public inputs satisfy the proof.

  • Compiling verifier contract:

    Head over to the Ethereum Remix IDE and paste in the smart contract.

    • Compile the smart contract using the correct solidity compiler version.
  • Deploying verifier contract:

    Use the MetaMask extension and connect to a testnet. I chose the Rinkeby testnet for this example.

    • Select Injected Web3 as the environment.
    • Select the Verifier contract and click Deploy.

    Deployment transaction: https://rinkeby.etherscan.io/tx/0x6b7c99bc9b6972e11c3a2918d8fe00b40f0036ebf18d28918a20a80b8d306947

  • Verify proofs using the deployed contract:

    Use the generatecall helper like we did in a previous step and paste it in to the verifyProof input field under Deployed Contracts:

    This is a pure function meaning it doesn't perform any writes. You'll get back a boolean response as you can see in the bottom right below the method input field.

Additional

  • To show general statistics of a circuit, you can run:

    snarkjs info --circuit=circuit.json
  • Print the constraints of the circuit by running:

    snarkjs printconstraints --circuit=circuit.json

Resources

License

MIT