Score:1

RSA accumulator

yt flag

In the ZeroCoin paper, it uses a zk-proof from Camenisch's dynamic accumulator that shows a Pedersen commitment hides an element of an RSA accumulator (https://link.springer.com/content/pdf/10.1007/3-540-45708-9_5.pdf). However, it looks like that the proof can also be used to prove that a subset of elements belong to the accumulator.

Now the question is: how does ZeroCoin prove that the Pedersen commitment hides a single element?

Score:2
gb flag

Every element in the accumulator is a prime, between some chosen bounds. In this case, the Pedersen commitment

$$ c = g^Sh^r \pmod{p}$$

is chosen so that $c$ is prime. Then, all these values $c$, each representing a "coin", are accumulated into the RSA accumulator $$u^{\prod {c_i}} \pmod{N}.$$

Whenever a coin is spent, the network checks that $c$ is indeed prime (meaning it cannot represent more than one accumulated value), and that its membership witness inside $A$ is valid. Note that this is done indirectly via the proof of knowledge you mentioned, but that proof of knowledge does indeed require $c$ to be prime. A proof is given in the full version of the paper, Theorem 4.

Sean avatar
yt flag
I was reading the ZeroCoin paper. For the coin $c$, when it's spent, it looks like it is NOT disclosed to the public (that's why Camenisch's zk-proof is used). The check of $c$ is prime is performed when $c$ is added (created) and saved to blockchain, but not when it is spent (because $c$ is anonymous). So here, when it's spent, I could actually try to submit a proof for $c_1 * c_2$ and submit proof $c_1*c_2 = g^{S_1 + S_2} h^r'$. This looks like an attack?
meshcollider avatar
gb flag
As mentioned, Theorem 4 of the full paper says that a valid proof also proves that $c$ is prime. You are correct that $c$ is not disclosed, so as to not leak where the spent coin came from. That's why I wrote "that this is done indirectly"
Sean avatar
yt flag
Thanks. I'd assume it's the range proof that limits it to be a single prime?
meshcollider avatar
gb flag
It is more than just a range proof, it is a full proof of knowledge, but that is just one of the things it proves.
I sit in a Tesla and translated this thread with Ai:

mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.