Blockchains operate as permissionless peer-to-peer networks, allowing any node to freely join or leave the network at any time. This dynamic nature introduces a significant churn rate, with nodes constantly joining and leaving the network. As a result, transaction propagation across this network is asynchronous, without any guaranteed transmission delays. The time it takes for a transaction to reach a particular node or miner is uncertain, as it depends on the ever-changing network topology and connectivity of the involved nodes.
To incentivize miners to include their transactions in the next block, users attach a transaction fee. Miners prioritize transactions with higher fees, essentially conducting a fee auction. Within each block, transactions are executed sequentially, and their order matters as it determines the resulting blockchain state for subsequent transactions. To ensure execution on the latest state, users must “frontrun” others by paying the highest fee, positioning their transaction at the very front of the next block.
Privacy is a major concern in these permission-less networks, as all transactions are publicly visible on the peer-to-peer network. Well-connected nodes with more network connections and better network performance have an inherent advantage, as they can potentially infer transaction sources and propagate transactions faster. Some actors exploit this by optimizing dedicated high-frequency blockchain clients, aiming to gain information asymmetry advantages over others.
Two prevalent fee auction designs exist: the price gas auction and the sealed bid gas auction. In the price gas auction, transactions propagate across the public peer-to-peer network, and miners prioritize higher fee transactions. In the sealed bid gas auction, traders send transactions to a centralized relayer, and only the winning bid is forwarded to the miner, making it a private auction. The chosen auction design impacts transaction inclusion strategies and the level of privacy afforded.
The public visibility of transactions, coupled with the potential for well-connected nodes to infer transaction origins through network analysis, exacerbates privacy concerns. The inherent information asymmetry favors actors with better network resources and optimized clients, potentially leading to unfair advantages in transaction propagation and inclusion.