Weighted Inner Product Based Range Proofs #21
Conversation
…dded range-proof benches to assess faster verification.
| hi_tag: &[GE], | ||
| g: &GE, | ||
| h: &GE, | ||
| P: &GE, |
There was a problem hiding this comment.
the prover is not using P explicitly, I wonder if we should input P to the prover and maybe sanity check the witness against it ?
There was a problem hiding this comment.
I think instead of passing P as a paramter to the prover function, it would be a better idea to let the prove function compute P using the witnesses. What do you suggest?
| use Errors::{self, RangeProofError}; | ||
|
|
||
| #[derive(Clone, Debug, Serialize, Deserialize)] | ||
| pub struct RangeProofWIP { |
There was a problem hiding this comment.
consider adding a statement struct that includes all the g_vec,h_vec,G,H etc... This way the verifier takes a proof and a statement and produce 0/1 .
There was a problem hiding this comment.
Done. I have not added the ped_com (commitments to amounts) in the struct as they need to be generated only by prover using secrets. I hope that is okay.
There was a problem hiding this comment.
What I am usually doing is defining a struct for Witness struct for Statement and struct for Proof.
prover gets as input Witness and Statement and output a Proof
verifier : is either a function of the Proof . Or gets Statement and Proof and outputs Result
Implementing range proofs with size 15% shorter than that of the original Bulletproofs protocol using weighted inner product argument introduced in Bulletproofs+ paper.
Write weighted inner product (wip) prover and verifier functions.
Fasten verification using multi-exponentiation in wip.
Scalar power and multiplication optimisations in wip.
Range proof using wip protocol and comparison with Bulletproofs.
Benchmark wip and the associated range proof.