4D Record | Zhang Ren is a guest on Zero Knowledge podcast, talks about consensus protocol security and NC-Max

0

 51 total views

Zero Knowledge is a well-known English podcast in the industry that focuses on zero-knowledge proofs, in-depth discussion of emerging Web 3 technologies, and community members who are building it. The audience of this podcast is fellow developers and people who want to understand the inner workings of the blockchain. The creators of the podcast, Fredrik Harrysson and Anna Rose, said that they do not speculate, talk about transactions, and focus more on the technology itself.

The most recent podcast invited Nervos researcher Zhang Ren. Zhang Ren and the host talked about his research on consensus algorithms, and started with Ren’s previous research on evaluating the security of the PoW consensus protocol, and discussed his NC-Max consensus protocol designed for Nervos. The Nervos Chinese community has compiled the podcast’s content for everyone to view.

Collection | Kazafa

Podcast creator Anna: The topic of consensus

Podcast creator Fredrik:

Ren:

I have been engaged in Bitcoin and cryptocurrency research since 2014,

Anna:

Ren:

Fredrik:

Ren:

Fredrik:

Ren:

Anna:

Ren:

For example, either you can prove that under the new consensus protocol, no matter what attack strategy the attacker adopts, he will not be able to make more profits or cause more losses; or you try to analyze all possible strategies the attacker may adopt, regardless of What the attacker does in a certain strategic space, he cannot cause more losses.

Fredrik:

Ren:

Anna:

Ren: Uniform tie-breaking protocol Rootstock FruitChains

Markov Decision Process

Anna:

Ren:

Anna:

Ren:

Anna:

Ren: Its Chain Quality is not perfect

Anna:

Ren:

READ  A review of Kucoin security incidents: Industry linkages are blocking and anti-hacking, hackers or loosing water?

Anna:

Ren:

“Better Chain Quality Agreement”

“Anti-attack protocol”

Fredrik:

Ren:

If I always mine on the latest chain with the smallest hash value, all honest miners will naturally mine on the same chain. The computing power will never be split and the chain becomes more secure.

Fredrik:

Ren: The performance of the minimum hash value splitting protocol is slightly better than that of the Nakamoto consensus. “Late to catch up”

Suppose you have a secret block and the hash value of this block is very small.

Anna:

Ren:

Anna:

Ren:

Fredrik:

Ren: “Subchains”

“Reward Lucky Agreement”

Whenever an attacker finds a weak block, it will detain a block, hoping that honest miners can find an official block so that I can use my weak block to invalidate the official block mined by honest miners

Fredrik:

Ren:

Anna:

Ren:

The first point is about how to analyze security.

“Reward the bad or punish the good”

Anna:

Ren:

The Markov Decision Process methodology has been extended

Anna:

Ren:

There is no difficulty in applying Markov decision process or deep learning to analyze the PoS protocol, but the problem with analyzing the PoS protocol is that you cannot directly aggregate a set of protocols and let them compete with each other, knowing who is the first and who is the second. There is no such simple result.

Anna:

Ren:

Fredrik:

Ren:

“Nakamoto consensus” (Nakamoto consensus, NC)

It has only a set of minimum security assumptions

It minimizes communication overhead

It is chain-based, which means that all transactions have a global order, and it does not have the problem of repeated packaging like the DAG protocol.

READ  Bitcoin sets a new record: A new ATH at $25k

Anna:

Ren: throughput delay

When I interned with Peter Wuille and Gregory Maxwell at Blockstream, we discussed a lot about how Bitcoin propagates blocks and transactions, which ultimately led to the birth of this research.

Anna:

Ren:

Fredrik:

Ren:

As long as we manage to find a balance point, that is, the block interval is low enough, the block is large enough, and all the node bandwidth is exhausted, then we have completed the task.

The definition of “new transaction” refers to a transaction that is included in the block before it has completed its propagation.

Anna:

Ren:

First, we decouple transaction synchronization and transaction confirmation.

The transaction must first appear in the transaction proposal area, and only after two blocks can it be embedded in the transaction commitment area.

Fredrik:

Ren:

Fredrik:

Ren:

Concurrent operation.

Anna:

Ren:

Anna:

Ren: Prism is another example of decoupling these two things

The transaction promise area and the transaction proposal area are re-coupled in the same block structure, which greatly simplifies the security analysis, because NC-Max still follows the Bitcoin backbone protocol and uses the “longest chain” principle.

Anna:

Ren:

“Transaction withholding attack” You can make a block full of new transactions. Your block actually spreads more slowly than all other blocks. This gives you some time to conduct selfish mining, but no one Can blame you

Anna:

Ren:

Anna:

Ren:

Anna:

Ren:

Nervos CKB uses UTXO model

Anna:

Ren:

Anna:

Ren:

lock script type script

Economic model, the native token CKB represents storage space

Anna:

Fredrik:

Ren:

READ  OKEx Talks 2019 (Valencia)- DAOS: Why Organizations Will Operate on Blockchain in the Future

You can use any type of asset to pay transaction fees, and there is no Gas mechanism

RISC-V to create a virtual machine

Fredrik:

Ren:

Fredrik:

Ren:

Anna:

Ren: “Lay down the Common Metrics 2.0”