Introduction to Saito Consensus

This is an introduction to Saito Consensus for Bitcoiners. It aims to give those familiar with proof-of-work a high-level overview of how Saito works. There is a also very short one-page summary of the entire mechanism on the Saito Wiki. Once you understand the basics, please see the Saito whitepaper for details on how Saito addresses issues not discussed here.

Saito is a layer-one blockchain that requires its blocks to contain a certain amount of “routing work”. Routing work is derived from transactions fees so in practice this means they need to contain a certain amount of money. How much depends on how long it has been since the last block and how many hops the transactions containing these fees have taken to reach the block producer.

Routing work punishes participants in long-chains. Transactions that have travelled only a few hops are more valuable than transactions that have travelled many hops. Transactions with shorter routing paths are more valuable for making blocks. And their routing nodes ultimately get paid more too. So what “routing work” is really measuring is how efficient nodes are at collecting transactions and getting them into blocks.

So in Saito nodes collect transactions and produce blocks once they have enough “routing work” in their mempools to meet difficulty criteria. The uneven flow of fees into the network ensures that someone will always be first to produce a block and creates a natural round-robin dynamic that boosts scaling and deters network centralization (the node that produced the last block is the least likely to to produce the next).

The critical step Saito takes to make this work is that when a block is produced the network burns all of the fees in that block. No-one is paid for producing a block when that block is made.

By not paying for block production, Saito guarantees that producing a block is always expensive. It is only — now — after Saito has produced a block that the network holds a payment lottery. Miners begin to hash in the hopes of finding a solution to a mining puzzle attached to the new block. If no solution is found our fees stay burned. But if a solution is found and included in the very next block our fees are unburned and split between the miner that found the solution and a lucky routing node. Mining nodes win proportionally to their hashpower. Routing nodes win proportionally to the amount of routing work they contributed to the block in question.

Because routing work degrades as transactions propagate through the network this system gives the user-facing nodes in the network the highest chance of getting paid. But as the last node in the routiung path of all transactions in the block, block producers do still collect payments on occasion.

Saito Consensus thus pays honest nodes a fraction of the fees they route into the network. But attackers? Attackers must burn their own money to produce a competitive chain. And attackers must then burn hashpower to get a fraction of their money back. Even if attackers are willing to burn money with no return, producing long chains is still impossible without supporting hashpower. But every dollar spent hashing earns them far less than a dollar back in fees. The math works out so the cost of attacking the network is always greater than 100 percent of the fee throughput of the honest chain. This is double the security of POW and POS, where the cost-of-attack is capped at 50% of inbound fee flow.

Saito thus generates twice the security of POW and POS networks, while paying for its network infrastructure instead of requiring additional and separate fees to cover those costs. This makes Saito at least 4x cheaper than competing mechanisms on a per-byte basis, with its cost savings growing as the network does.

Saito Consensus has mechanisms that provide protection against spamming / block-flooding attacks, and that make sybil attacks unprofitable. In addition, Saito introduces new defensive mechanisms that are not possible in other chains, such as weaponizing routing so that honest nodes can help defend the network against attack simply by denying transaction-flow to attackers. We encourage readers interested in these next-generation mechanism to dig deeper into the Saito whitepaper.

The improvements that Saito offers are fundamental. As unlike other blockchains which cannot differentiate between honest nodes and attackers, Saito leverages the fact that attackers are always +1 hop from an honest node or fee-paying user. Saito adjusts the cost of producing blocks up and the income from producing blocks down for nodes that are deeper in the network. Interior nodes must find strategic positions where multiple streams of inbound user-fees flow into the network to profit.

Once the overall mechanism is understood, what is striking about Saito is the simplicity of its approach and the elegance with which it removes the economic attacks in other chains that are created by reliance on outside markets (for hashpower or stake) and market failures within the consensus mechanism.

We encourage everyone interested in decentralized cryptocurrencies to learn how Saito works and join us on Saito Telegram. Questions, feedback and suggestions are welcome anytime.