We are looking into switching into a DAG - I have a feeling this is the way to go here - We'd like to test a thing though
Switching to a DAG instead of a Blockchain
I have a feeling HON can improve using the starting point of IOTA's DAG protocol and perhaps even remove the necessity of the coordinator with the substitution of two previous nodes validation by Markov Blankets validations along with Merkle Trees
The Idea
Let me elaborate. A Markov Blanket is a set of information that renders a group of nodes independent of the rest. This is composed by the parents, the children and the children's parents.
Therefore, an incoming transaction would verify the children's Markov Blanket in order to provide a reproducible verification unit by a simple Merkle tree in the Poisson distribution related to the validation round.
The Problem we'll be working afterwards
Calculate the probability that a double-spend will be detected once Markov Blanket Merkle Trees are randomly verified by the network as in opposition to 2-previous transactions.
References
- http://www2.im.uj.edu.pl/katedry/seminaria/inda/wp-content/uploads/2009.11.30.kulaga_tomasz.pdf
- https://eprint.iacr.org/2013/881.pdf
Working on that right now. Cheers...
EDUARDO CAPANEMA
@bitworkers