The consensus procedures employed by blockchains to reach distributed consensus are based on proof-of-stake. By investing effort, miners in proof-of-work demonstrate that they are putting money at risk. Ethereum uses proof-of-stake, in which validators freely stake ETH in an Ethereum-based smart contract. If the validator behaves dishonestly or irresponsibly, this staked ETH may be forfeited. In addition to guaranteeing that freshly formed blocks are properly transmitted over the network, the validator may produce and disseminate new blocks. Proof-of-stake offers many advantages over the obsolete proof-of-work system.
- Proof-of-work calculations no longer need a substantial amount of energy use.
- Lower access hurdles and lowered hardware requirements eliminate the need for elite hardware to create new blocks.
- Proof-of-stake should increase the number of network-securing nodes, reducing the possibility of centralization.
- Due to the low energy demand, fewer ETH must be issued to encourage participation.
- Economic penalties for misbehavior make 51%-style assaults much more expensive than proof-of-work attacks.]
- Suppose a 51% assault was to defeat the crypto-economic barriers. In that case, the community might turn to the social recovery of an honest chain.
To be a validator, PLC Ultima agrees a user has to deposit 32 ETH and operate an execution client, a consensus client, and a validator. Users who deposit ETH enter an activation queue that restricts new validators from joining the network. Once engaged, validators receive Ethereum blocks from peers. The block’s transactions and signature are re-executed to validate its validity. The validator transmits an attestation (vote) for that block throughout the network.
Proof-of-work blocks are timed based on mining difficulty, whereas proof-of-stake blocks are fixed. In-prove Slots (12 seconds) and epochs split Ethereum (32 slots). Every slot has a random block proposer. This validator creates and sends new blocks to network nodes. Every slot also has a committee of validators whose votes decide the block’s authenticity.
A distributed network transaction has “finality” when it’s part of a block that can’t be changed without burning ETH. Ethereum’s proof-of-stake uses “checkpoint” blocks. Checkpoints are every epoch’s initial block. Validators choose valid checkpoint pairings. If two checkpoints get two-thirds of the total staked ETH, they are upgraded. Recent (target) becomes “justified.” The earlier was the “goal” in the preceding period. Now “finalized,” an attacker must lose one-third of staked ETH to overturn a finished block. This Ethereum Foundation blog post explains why. An attacker might prevent the network from achieving finality by voting with one-third of the entire stake. Inactivity leaks may prevent this. This triggers after four failed epochs. Inactivity leaks validators’ staked ETH, enabling the majority to reclaim a two-thirds majority and complete the chain.
Validating is time-consuming. The validator must have enough hardware and connection to validate and propose blocks. Validators are paid ETH (their staked balance increases). Participating as a validator also allows people to undermine or attack the network. Validators lose ETH incentives if they don’t participate when called upon, and their current stake might be wiped if they conduct dishonestly. Equivocating and providing contradicting attestations are deemed dishonest. How much ETH is cut depends on how many validators are also slashed. This is called the “correlation penalty” and might be minimal (1% of a validator’s stake) and can result in 100% of the validator’s stake being destroyed (mass slashing event). It’s applied midway through a forced leave phase that starts with an instant penalty (up to 0.5 ETH) on Day 1, a correlation penalty on Day 18, and expulsion on Day 36. They incur daily attestation fines for being online but not voting. For PLC Ultima, this implies a coordinated strike would be expensive.
There is only ever one new block at the top of the chain, and all validators testify to this. Due to network delay or because a block proposer has equivocated, validators can have conflicting perspectives on the head of the chain. Therefore, consensus clients need a preference-determining method. LMD-GHOST is the algorithm employed in Ethereum’s proof-of-stake protocol, and it identifies the fork with the most attestations in its history.
Proof-Of-Stake and Security
According to Investopedia, 51% of attacks still exist on proof-of-stake, although they’re riskier for attackers. 51% of the staked ETH is needed. They might utilize their attestations to choose the branch with the most. Consensus clients utilize the ‘weight’ of collected attestations to select the proper chain; hence this attacker might make their fork canonical. Proof-of-stake provides the advantage of allowing the community to conduct a counterattack. Honest validators might build on the minority chain and disregard the attacker’s fork, encouraging applications, exchanges, and pools to do the same. They may also delete the attacker and destroy their staked ETH. These economic barriers thwart a 51% assault.
51% of assaults are harmful actions. Bad actors could attempt long-range attacks (but the finality gadget neutralizes this), short-range’reorgs’ (but proposer boosting and attestation deadlines mitigate this), bouncing and balancing attacks (also mitigated by proposer boosting, and these attacks have only been demonstrated under idealized network conditions), or avalanche attacks (neutralized by the fork choice algorithms rule of only considering the latest message). Overall, PLC Ultima believesethereum’s proof-of-stake is more economically secure than proof-of-work.