How Gee Consensus Works

The Gee Consensus protocol relies on a unique iterative subsampling voting process that involves gathering responses from a small, random sample of validators to reach agreement. Here’s how it works:

1. Sampling and Validation:

Ÿ When a node receives a new transaction to validate, it polls a random subset of validator nodes in the network, asking each for its stance on the transaction.

Ÿ Each validator evaluates the transaction and responds, indicating whether it believes the transaction should be accepted.

2. Iterative Voting Process:

Ÿ If a majority of the sampled validators (determined by a parameter called α, or alpha) prefer a particular transaction, this majority vote influences the node’s decision to support that transaction as well.

Ÿ Nodes continue this process, repeatedly sampling validators and adjusting preferences based on majority responses.

3. Reaching Consensus:

Ÿ When validators produce the same result across a sufficient number of rounds (defined by a parameter called β, or beta), consensus is achieved. At this point, all honest validators align on the decision to accept or reject the transaction.

Ÿ Both α (majority threshold) and β (consecutive agreement threshold) are configurable parameters, making the system adaptable to network size and transaction types.

4. Handling Conflicts:

Ÿ If conflicting transactions arise, validators enter a feedback loop that allows them to quickly converge on the correct, non-conflicting transaction. Honest validators ultimately rally around a single transaction, ensuring network consistency.

5. Invalid Transactions:

Ÿ Transactions that do not meet validity requirements (e.g., a request to send more funds than available) are automatically excluded from the consensus process, preserving the network’s integrity.