When Alice Goes Bad: Byzantine Secret Sharing
We need to start building systems which are fault-tolerant, and where we assume that sometimes our systems give us errors, or can be faulty, or that they have been taken over by a malicious agent. This will give us Byzantine Fault Tolerance (BFT) in our processing and decision-making. For example, let’s say we are processing a transaction, and have four checkers for the validity of the transaction. If we had one checker, then it may have an error or could be compromised by a hacker. But if we have four, then if three of the checkers were good, we would have an election, and take the majority vote. A checker which loses these elections may then be faulty or is compromised.
A perfect way of keeping things secure and creating resilience is to use Shamir Secret Sharing (SSS), and where we can distribute a secret, and then allow any n-from-m to recover the secret. In this way, Bob, Alice, Carol and Dave can share a secret, and then any three of them can come together and recreate the share. This means we also have resilience in that one might not want to share — or have lost their secret — and we will still be able to recover it. But, what happens when Alice goes bad, can we recreate the shares between Bob, Carol and Dave, so that Alice cannot rebuild the secret? Well, yes! For this, we use Proactive Secret Shares (PSS), but first I’ll explain Shamir’s Secret Shares…
