Lec 7-2: Proof-of-Stake

Proof-of-Stake

What is PoS?

Proof-of-stake is an alternative to Proof-of-Work

• PoW might lead to a low network security with block incentives that decline over time -> miner ⬇️
• PoS is one way of changing the miner's incentives for higher network security

In PoS based public blockchain, a set of validators take turns proposing and voting on the next block, and the weight of each validator's vote depends on the size of its deposit (=stake)

• Someone holding 1% of coins can mine 1% of the "Proof of Stake blocks"
• Anyone can become a validator by staking (= locking up their coins into a deposit)
• All current validators can participate in a consensus algorithm

 

Benefit

• No need to consume large quantities of electiricity for security
• Don't need to issue as many new coins in order to motivate parcipants to keep participating in the network
  • negative net issuance is possible theoretically: a portion of transaction fees is burned and so the supply goes down over time
• Descourage centralized cartels, Reduce centralization risks
• Ability to use economic penalties to make various forms of 51% attacks vastly more expensive to carry out than PoW

---

Chain-based PoS

Chain-based PoS

• pseudo-randomly selects a validator during each time slot
• assings that validator the right to create a single block, and this block must point to a previous block(normally the block at the end of the previously longest chain)
• Therefore, over time most blocks converge into a single constantly growing chain.

Nothing-at-stake Problem

a scenario where block creators on generic proof-of-stake protocols have nothing to lose when the network forks.

• If there are only rewards for producing blocks and no penalties -> In the case that there are multiple competing chains, validators try to make blocks on top of every chain at once for incentives

 

Voting for both without any penalty

• If all actors are economically rational(acting for maximizing benefit), then a blockchain may never reach consensus even if there are no attackers.
• Attacker needs only overpower altruistic nodes and not rational node(who would stake on both the original chain and the attacker's chain)

 

Solution for Nothing-at-stake Problem

Penalizing validators if they simultaneously create blocks on multiple chains

 

Punishing validators for creating blocks on the wrong chain

 

 

These two solutions can be combined together

• Those solutions can replicate the economics of PoW inside of PoS
• In PoW, there is a penalty for creating a block on the wrong chain that is implicit in the external environment
  • miners have to spend extra electricity and obtain or rent extra hardware
• Using those solutions, in PoS, it makes explicit penalties
  • it imposes slightly more risk on validators (although the effect should be smoothed out over time)

---

BFT(Byzantine fault tolerance)-style PoS

BFT-style PoS

• Validators are randomly assigned the right to propose blocks, but agreeing on which block is canonical is done through multi-round processes where every validator sends a vote for some specific block during each round
• All validators permanently agree on whether or not any given block is part of the chain.
• Consensus on a block can come within one block, and doesn't depend on the length or size of the chain after it.
• 블록 제안과 블록 선택 과정이 분리 -> 블록 제안 과정에서는 랜덤하게 선택된 여러 검증자들이 블록을 선택. 이후 블록 선택 과정에서 지분 만큼 투표를 하고 가장 많은 투표를 받은 블록이 다음 블록으로 선택됨

 

Byzantine General's Problem

• In group of generals where they must only decide whether to attack or retreat
• Every general agress on a common descision, for a half-hearted attack by a few generals would become a rout
• Presense of traitorous generals -> cause problem
  • EX) 9 generals are voting, four of whom suppport attacking while four others are supporting retreatiing, the ninth general may send a vote of retreat to those generals in favor of retreat, and a vote of attack to the rest.
  • If the generals are physically separated and using messengers who may fail to deliver votes or may forge false votes, the problem is further complicated

 

Byzantine fault

• Formally defined:
  • $n$ components, $t$ of which are dishonest, and assuming only point-to-point channels btw all components
  • whenever a component $A$ tries to broadcast a value $x$, other componenets are allowed to discuss with each other and verify the consistency of $x$, and eventually settle on a common value $y$
  • This system is said to resist Byzntine faults if a componenet $A$ can broadcast a value $x$ then,
    • If A is honest, then all honest components agree on the value $x$
    • In any case, all honest components agree on the same value $y$