On September 5th, the POD conference hosted by Odaily and co-organized by 36Kr Group was held in Beijing. At the conference, Li Xiaozhou, co-founder of Celer.Network, delivered a keynote speech on "Build and Operate Internet-Scale dApps on Celer Network".

Li Xiaozhou believes that the reason for the slow speed of the public chain is the irreconcilable contradiction between the consensus on the chain and the scalability in the distributed system. The improvement of the consensus algorithm can certainly improve the performance of the public chain, but it is still far from making the blockchain an Internet-scale application, so it must rely on the off-chain expansion system.

He said that the blockchain is considered to be Internet 3.0, but the current blockchain cannot even see the shadow of the Internet in terms of technology and practical application. If we want to build a globally shared distributed ecosystem, all computers, mobile terminals, and IoT devices in the 5G era can conduct large-scale, decentralized, safe and fast transactions on it. Li Xiaozhou believes that the performance that this system needs to provide is billions of calculations per second.

The root cause of the slowness of the current public chain is the irreconcilable contradiction between on-chain consensus and scalability in distributed systems. If you want all nodes in a distributed system to reach a consensus on each operation, even if the huge communication cost is ignored, the performance of the entire system will not exceed the slowest node in the system. The improvement of the consensus algorithm can certainly improve performance, but it is still far from realizing Internet-scale applications. Therefore, we must rely on the off-chain expansion system, that is, build a new layer of expansion platform on the underlying public chain to greatly improve the performance of distributed applications, so that the blockchain can be quickly implemented.

Off-chain expansion includes chain testing and state channels. When the two parties cannot reach an agreement, the on-chain consensus will be resorted to. Just like the law enforcement and judicial system in reality. In daily economic activities, there is no need for the judicial enforcement system to intervene in every transaction too frequently. You only need to know that the other party has a deposit, and there are contracts and evidence when conducting transactions, and you can go to court when disputes arise. . Off-chain expansion is also the same idea. The most precious resources of the underlying public chain are only used in operations that require consensus on the chain, such as deposit and withdrawal, arbitration enforcement, and so on. This can improve the performance of distributed applications and also protect the privacy of users.

For example, how to use off-chain expansion technology to make fast payments. If there are two people who don't trust each other and want to conduct frequent transaction interactions, they can first deploy a smart contract on the chain. The function of this smart contract is to help them manage their assets. After the initial on-chain conditions are set, the two can interact frequently off-chain. They only need to co-sign to update the state, including a serial number, indicating a timestamp, and indicating how much money on the chain should belong to whom. This operation is very fast, because it is only limited by the speed of each person's own processing, and the communication cost between the two.

At any time, any party can submit the state of the two people's co-signature held in their hands to the smart contract on the chain to get back their own money.

If the other party agrees to this state, or does not respond within a specified time, then the smart contract on the chain will distribute the money according to the state it receives.

Of course, any party can also determine that the other party has fraudulent behavior. For example, if he submits an overly factual evidence against himself, and he has the latest signature status of two people, he can submit this status to the chain, so that Arbitration on the chain ensures that the interests of the trustworthy party are not harmed. Through such a setting, two people who do not trust each other can conduct transactions very quickly.

In a blockchain ecosystem, various accounts can be connected into a huge network through state channels. In this network, any two nodes can make fast off-chain payments through direct or multi-hop forwarding. Such a system is like the current Internet, as long as there is a network cable, transactions can occur with any node. It can bring many opportunities for global ultra-high-speed P2P payment, and can give birth to many new business models.

The above fast payment function is just one of the application examples of off-chain expansion. To further expand, when you don't need mutual trust, you can conduct all kinds of interactions with complex logic, and only when you feel it is necessary, such as when one party fails to cooperate, will you resort to consensus on the chain Let it arbitrate, so the overall cost will be low.

Hello everyone!

The following is the full text of the speech:

Hello everyone!

Many people refer to the blockchain as Web3.0, and its corresponding Internet 2.0 realizes the automatic transmission of information on a global scale, and the blockchain realizes the automatic transmission of value on a global scale. What difference do we see between the blockchain of 3.0 and the Internet of 2.0?

On the most popular mainstream blockchain at present, only about ten calculations per second can be performed. It is obvious that the blockchain cannot even see the shadow of the Internet in terms of technology and practical application.

How fast does a blockchain need to reach the general popularity of the Internet? If our goal is to build a globally shared distributed ecosystem, so that all computers, mobile terminals, and IoT devices in the 5G era can conduct large-scale, decentralized, safe and fast transactions on it. We think that the performance that this system needs to provide is billions of operations per second.

When you see this number, you may feel incredible, because it is too far from reality. Today I would like to share with you whether we can and how we can achieve such an ambitious goal. Most of the distributed applications now are mainly smart contracts written directly on the underlying public chain. We also know that the speed of the current underlying public chain is often very slow. Although it still has a lot of room for improvement, its development is also full of difficulties. The root cause of the correction is the irreconcilable contradiction between the on-chain consensus and scalability of the public chain in the distributed system. If we want all nodes in a distributed system to reach a consensus on each operation, even if these very huge communication costs are ignored, the performance of the entire system will not exceed the slowest node in the system.

There are various methods, including the improvement of the consensus algorithm, which can improve the performance of the public chain, but if it is to achieve Internet-scale applications, there is still a long way to go.

Therefore, we believe that we must rely on the off-chain expansion system, that is, build a layer of expansion platform on the underlying public chain to greatly improve the performance of distributed applications, so that the blockchain can be implemented quickly. What exactly is off-chain scaling technology? Some basic concepts of off-chain expansion, such as chain testing, such as state channels. For example, when the two parties cannot reach an agreement, they will resort to the consensus on the chain.

Here we can compare the blockchain system with the economic activities of real daily life. The underlying public chain is authoritative and fair, but its processing speed is relatively slow, just like the law enforcement and judicial system in reality . In our daily economic activities, there is no need for the judicial law enforcement system to frequently intervene. We only need to know that the other party has a deposit, and there are contracts and evidence when conducting transactions, and it is enough to be able to file a lawsuit when a dispute arises.

That is to say, the most precious resources of the underlying public chain are only applied to operations that require consensus on the chain, such as the deposit and withdrawal, such as arbitration enforcement, etc. This can not only greatly improve the performance of distributed applications, but also protect the privacy of users, because only the participants in the transaction know the details of these transactions when everyone cooperates.

Let's take a simple example below, how to use off-chain expansion technology to make fast payments? For example, there are two people who do not trust each other. If they want to conduct frequent transaction interactions, they can first deploy a smart contract on the chain. The function of this smart contract is to help them manage their assets. At the same time, according to the status of the two co-signatures, the custody assets are redistributed. When these initial on-chain conditions are set, the two can interact frequently off-chain. The specific method is that they jointly sign to update the state, which has a serial number, indicates a timestamp, and indicates how much money in the chain should belong to whom. This operation is very fast, because it is only limited by the speed of each person's own processing, and the communication cost between the two.

At any time, any party can choose to submit the state of the two co-signatures held in their hands to the smart contract on the chain to get their own money back. If the other party agrees to this state, or does not respond within a specified time, then the smart contract on the chain will distribute the money according to the state it receives.

Of course, any party can also determine that the other party has fraudulent behavior. For example, if he submits a false evidence that is not good for him, and he has the latest signature status of the two people in his hand, he can also submit this status to the chain. Let the chain Arbitration. Therefore, it can ensure that the interests of the honest and trustworthy party will never be damaged. Through this setup, two people who do not trust each other can conduct transactions very quickly.

In a blockchain ecosystem, any two nodes can make fast off-chain payments. This system is like our current Internet. You can travel to any corner of the Internet with a network cable at home. It can It brings many opportunities for global ultra-high-speed P2P payment, and can give birth to many new business models.

Let's look at a more general example. As I said just now, the core principle of off-chain expansion is that two people jointly maintain a multi-signature state. This state can be simple on-chain who owns how much of these deposits, or any code contract state.

For example, here are two people who want to play a chess game, and they can jointly maintain the code and board state of playing chess under a chain. And at the beginning, each gave the other party a conditional payment, for example, if the other party won the game within a certain period of time, they would give the other party a certain bonus.

Then the two people can update the code state that only they have by co-signing to play a quick game. The whole process is completely private, and when one party wins, if the losing party cooperates, he can choose to pay directly under the chain and clear the previous conditional payment.

Here we can see that when the two parties to the transaction are fully cooperative, the entire chessboard state from beginning to end, including its code and conditional payment, is completely on the public chain, without any operation calculation or storage, then it The cost will be very low and the operation speed will be very fast.

Of course, when one party does not cooperate, the party whose interests are damaged, because he has all the states signed by the other party, he can choose to submit this state to the chain for arbitration on the chain, as well as the previously sent to him by the other party. The record of conditional payment, so that the smart contract on the chain will analyze this state to see if the condition is met? If it is satisfied, the assets on the chain will be forcibly allocated.

So, just like before, we can see that with such a mechanism, when people don’t need to trust, they can carry out all kinds of very complicated transactions, and only when they think it is necessary, for example, if one party fails to cooperate Only when it is necessary to resort to consensus on the chain, so the overall cost is very low.

Next, I would like to briefly introduce Celer.Network’s off-chain expansion work. Celer.Network is based on off-chain expansion technology, a distributed platform that can be built on an Internet scale and extremely protects privacy. Its purpose is to allow everyone to quickly and easily develop and run distributed applications on the platform, as well as use massive scalability. Celer.Network itself is not an independent public chain, its architecture is mainly divided into two parts:

1. A complete set of off-chain expansion technology.

2. The economic mechanism of off-chain expansion.

Let me briefly introduce these different parts. Celer.Network will have formal verification to ensure its security, and can run on various public chains. Here we further extend the example of two people playing chess just now. Suppose there is a state channel between A and B, and a state channel between B and C. At this time, A and C can communicate through B. Play chess games under the chain. And we are like him at the same time, between ABC, the two complement each other and need trust.

Then on top of Celer.Network is the routing algorithm of the entire payment network, which is the core that enables this blockchain to form an ecological network. First of all, it has very high performance and can provide dozens of times the throughput of the current algorithm. It is completely decentralized, can be easily attacked by malicious nodes, and can greatly protect user privacy, allowing other nodes including intermediary forwarding nodes to There is no way of knowing who the sender and recipient of the money are.

Further up is our COS, the development framework and operating environment of the entire distributed application. It can provide a general distributed application design framework and interface, and can connect the state of the chain and the chain, and track the state of the chain. It also handles off-chain disputes, supports high-performance concurrent operations, and allows a distributed application to run on various public chains very easily.

Having said so many advantages of off-chain expansion, of course we must also emphasize that there is no free lunch in the world. While off-chain expansion greatly improves the performance of blockchain distributed applications, it also naturally brings some new challenges and new challenges. The problem. At this time, a new economic mechanism is used to improve the entire ecological environment.

The first challenge we have to deal with is the availability of status. If a user's status is guaranteed, it can deal with all malicious arbitration requests that may be made by the other party in real time. Celer.Network has designed a state guarding network. Users can choose to synchronize the state with the user protection network. If arbitration occurs, the state guarding network can submit the state on behalf of the user. The Celer economic mechanism can ensure that each user does not need to trust the nodes in the state guard network.

Another challenge to deal with is the security deposit. If I want to run an off-chain payment service and want to earn transfer fees, my first challenge is to have enough security deposits on each state channel. The Celer economic mechanism can encourage everyone to use their idle funds in the ecological environment of the chain, and can enable technical off-chain service providers to obtain these deposits through crowdfunding, while ensuring 100% security for all users.

As a summary, Celer.Network greatly improves the performance of distributed applications through a set of overall off-chain expansion technologies and innovations in economic mechanisms. Our goal is to bring the scale of the Internet into each blockchain. thank you all!