It's not bad idea to cut Ring Signature from calculating TX ID, but do you think it's possible to combine Schnorr Signature with Ring Signature to reduce overall input/decoy size in transaction?
Hi ETFbitcoin,
That is an interesting concept, combining Schnorr Signature with ring signatures is possible although it would require further study to observe the cost benefit analysis. Herranz et al. proposed such a system in their 2003 ePrint:
"Forking Lemmas in the Ring Signatures' Scenario" however that system is still linear.
Hi ETFbitcoin,
Upon further study: a transaction in our system employs (linkable) ring signatures in two different places. On one side, the sender signs his transaction with the linkable ring signature scheme. On the other side, a ring signature is required in the process of proof-of-sum. If one of them is replaced by Schnorr's signature, the anonymity of the whole system will be decreased. Moreover, Boolberry uses a
special flag [2] to guarantees that a coin won't be spent without mixins. Such a technique is to deal with the problem caused by signing a transaction on behalf of a ring with only one participant. Since the Schnorr's signature can be regarded as a ring signature with one participant, we don't recommend combining it with ring signatures in Boolberry v2.
[1] -
https://eprint.iacr.org/2017/921[2] -
https://www.slideshare.net/boolberry/boolberry-solves-cryptonoteflaws-37055246That's unexpected since Schnorr Signature can increase privacy when it's combined with non-privacy coins (everyone can see sender,receiver and the amount) when there's more an input, but thanks for the info.