[Cryptography] A new, more efficient consensus protocol

[Vincent Strom]

Dear Greg,

Certainly, I would be happy to elaborate. The main idea is that transactions work as lottery tickets. Not all transactions, but only those which have some min transaction fee and whose hash is less than certain threshold. We adjust this threshold so that the lottery rate is 10/s. Because the definition of lottery is not user specific, anyone can filter transactions for lotteries and maintain a separate mempool for them.

If this is so, then an active user only has to check if she is winning any of the available lotteries. For this she needs a hashrate of 10H/s (lottery rate). If a separate mempool is not maintained because, say no one on the network is an altruist, then the user has to check all the transactions. Firstly to see whether they are valid tickets and secondly if she is winning the ones that are. For this she would require a hashrate of ~2KH/s (assuming this is the transaction generation rate).

A hashrate of more than this is not useful because there is nothing more left to check. To get an unfair advantage, the powerful user may flood the network with transactions so that she could utilize her high hashrate but she is disincentivized to so due to the requirement of a min transaction fee as a qualification for being a lottery ticket.

Do let me know if something is still unclear.

Regards,
Vincent
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.metzdowd.com/pipermail/cryptography/attachments/20210115/294c0177/attachment.htm>