Author: & have spent Ittay Eyal
Introduction to selfish mining in | blockchain
Source: ethereum fans
Translation: & have spent O sword
The PoW block chain implements a form of State Machine Replication (SMR). Unlike traditional SMR protocols, the PoW blockchain is open, meaning anyone can join in, and the system provides financial incentives for participants (also known as "miners") to comply. Therefore, what is also different from the traditional SMR protocol is that when reasoning about the security of blockchain, simply assuming the number of malicious participants is often not the answer. The key is to ask whether the miners really have enough incentive to abide by their agreements. This is what this article is about.
To make the discussion more specific, let's limit it to nakamoto's bitcoin protocol. Ling has provided some background, as well as a security analysis of malicious adversaries. In our analysis, we intend to describe the system as a game between miners.
The game
Gamers are miners. The game is played in rounds, in which one miner can publish a block and the other miners can publish the block. Also, the in-game message is synchronous, so all miners receive the block from the previous round.
This, of course, simplifies the reality. For example, the model ignores the slow changes in the mining forces in the system, as well as the occasional, though rare, occurrence of block-out collisions (or bifurcations). Nevertheless, this model is sufficient as a first order approximation.
The deal was for each miner to block the longest chain, or, if the two chains in the fork were the same length, to follow the one they had received first.
Each player in the game is committed to maximizing his or her own revenue -- this is his or her utility function. Specifically, we also assume that this is an infinite-horizon game, that is, as the playing time approaches infinity, a miner's return is its average payout ratio. This means that the reward in cryptographic currency is given to a miner according to the block he has given. Note that blocks outside the main chain do not enter the miner's earnings.
Important to observe
The proportion of outages on the main chain is a real-time simulation of the miner's return.
Assume that the mining force in the system is static, the system every 10 minutes out of a block, attack immediately after a difficulty adjustment. Suppose a blocking strategy causes a certain percentage of the blocks in the network to be discarded. For example, 20% of all miners' blocks will be generated outside the main chain, and this proportion is stable. So, while the system is still producing one block every 10 minutes, only 80% of it is going to be on the backbone, which means the backbone is going to grow at a rate of one block every 12.5 minutes instead of one block every 10 minutes. The difficulty of the bitcoin protocol is adjusted once for every 2,016 blocks, so it takes longer than usual to adjust the difficulty (12.5×2016 minutes instead of 10×2016 minutes).
As soon as the difficulty adjustment occurs, the difficulty drops again, bringing the block interval back to 10 minutes. This means that the system as a whole is getting out of the block faster, one block every eight minutes.
Therefore, if a miner has an equal proportion of the net, and the proportion of the blocks on the main chain is binary '. , the return per hour will be proportional to '.
Selfish mining algorithms
Selfish Mining is a speculative Mining algorithm used to prove that the agreement is not an equilibrium for small miners. Let's take a look at the mechanism of selfish mining, and then discuss why and when it works.
At first, the selfish miners would dig up the longest chain, as the agreement intended. However, once they have dug up a block, they hide the block rather than publish it immediately, and then try to continue after the secret block to form a "secret branch".

At the same time, other miners extend the open chain, which will eventually get longer (with a probability of 1) because they account for most of the mining effort. And the selfish miners continue to prolong their secret branch until the open one is one block behind. Then the selfish miner would publish his secret branch.
& have spent

& have spent
Since the secret branch is longer, the other side will think that this is the main chain, and from that point on, everyone will follow the selfish miner's branch, while the blocks dug out by the other miners will be discarded -- ignored -- leaving the miners with nothing.
& have spent
But this strategy is not foolproof -- selfish miners have been taking risks since they started digging in secret. If TA gives a secret block and the other miners give a block, TA can't release the secret block to become the longest chain; Instead, you end up with two equally long branches competing for the longest chain.
& have spent

& have spent
The selfish miners tried to extend their branches; To simplify the analysis, we assume that other miners will also try to extend their own branches. If the TA can beat the next block, the branch becomes the longest chain, and then the next attack is resumed at the end of the longest chain. If other miners are born, selfish miners are at a disadvantage (TA's chain is shorter). In this case, the TA will abandon the attack and look for the next opportunity. In this attack, his secret branch becomes a shorter branch, leaving him with nothing.
& have spent
Selfish mining analysis
& have spent
At first glance, this attack should not work - selfish miners are in the minority and must win more than they lose. But a careful analysis suggests this is not always the case. The game is naturally described as a Markov Chain. By calculating the selfish miner's block and the other miner's block, we can calculate the ratio of the selfish miner's block (and the benefit) to the main chain, which is a function of the force scale.
& have spent

We can see that selfish miners, who account for more than a third of the work force, can increase their profits by breaking their agreements and implementing selfish mining algorithms.
& have spent
conclusion
& have spent
On the surface of the above analysis, when the calculation force of selfish miners exceeds 1/3, the selfish mining strategy is more profitable than the honest mining strategy, but this is under optimistic assumptions. For deeper analysis (including weaker models and ways to strengthen protocols), see the paper in Financial Crypto 2013 and the paper at ACM 2018.
& have spent
Follow-up & have spent Studies, including a recent one, have used markov methods to determine force thresholds for the dominance of honest mining strategies (as used in this article).
& have spent
Thanks to Ittai Abraham for your informative feedback. Please send us some comments on Twitter.
(after)
& have spent
The original link: https://decentralizedthoughts.github.io/2020-02-26-selfish-mining/
Welcome to join the official BlockBeats community:
Telegram Subscription Group: https://t.me/theblockbeats
Telegram Discussion Group: https://t.me/BlockBeats_App
Official Twitter Account: https://twitter.com/BlockBeatsAsia