Provable Single Secret Leader Election

We suggest the use of an adaptation of the Bayer-Groth permutation argument here to obtain a secret single-leader election with low prover overhead. (Related post here.)

Construction
Let g be an elliptic curve generator. Let v_1, ...,v_k be a list of validators to secretly shuffle for block proposals. Every validator v_i has a permanent public key pk_i as part of their validator record where pk_i=g^{sk_i} for some secret key sk_i . To begin an ephemeral base is set to epk = g.

To shuffle a set of ciphertexts participate in the election a validator v broadcasts:

  • a new ephemeral base epk' = epk^r
  • shuffled public keys pk_1', ... , pk_k'
  • a corresponding SNARK proof \pi with public inputs pk_1,...,pk_k, epk and private inputs (r, \sigma ) such that pk_i'=pk_{\sigma(i)}^{r} and epk'=epk^r

A participant can identify their public key as the value pk_j' such that pk_j' = (epk')^{sk_i}. If the shuffle is accepted, then the ephemeral base is updated to epk' = epk and the public keys are updated to (pk_1, ... , pk_k) = (pk_1', ... , pk_k').

To limit the damage of a dishonest shuffler, it will be necessary to commit to the shuffle \sigma in advance of knowing the current ordering of public keys.

Motivation
Justin Drake proposed a low overhead secret single leader election. However, for security, his idea required the use of a private broadcast mechanism (e.g. Tor). Recently Dan Boneh, Saba Eskandarian, Lucjan Hanzlik, and Nicola Greco proposed a means to remove the private broadcast mechanism by instead encrypting the shuffled ciphertexts here. In this proposal we specify a means to instantiate the zero-knowledge shuffle argument.

For more technical detail see here

19 Likes

How viable would it be to have this included with the initial release of the merge?