Inspiration
There is no method for having consensus of time across a distributed network without a central entity
What it does
DCST leverages GPS atomic timing for periodic time referencing and accuracy maintenance. Zk proofs are then used to ensure no malicious acting. As long as the nodes time falls within the acceptable range, all it's actions are valid. This is done through consensus of a DA layer.
PLEASE SEE THE README FILE ON GITHUB
How we built it
Leveraged the output of current GPS protocols already established on devices. Created zk proofs to ensure that it's not tampered with. Then used this to update a locally run node clock at regular intervals. Created a basic gossip protocol to gossip time requests to nodes without regular updates. Then submit the proofs of timings to the DA layer for validity proofs.
Challenges we ran into
Time is impossible to realise a consensus on if participants rely on other actors for the truth as time will have passed. Additionally, having no single or a few sources of the truth means that there is no foundational basis for the truth and that is incompatible with creating an objective time truth.
Accomplishments that we're proud of
Coming up with an idea that flips consensus upside down. Here everyone has a source of the truth and that is validated by consensus, rather than a system with multiple parameters where no individual participant has a source of the whole truth.
What we learned
There are a lot of issues with latency and accuracy of time. Also time is a strange concept.
What's next for Distributed consensus of synchronised time
Build this into a mechanism for ensuring objectively fair FIFO block ordering. Effectively removing the possibility for MEV.
Log in or sign up for Devpost to join the conversation.