One change is that the current Internet (information network) will become a blockchain Internet (chain network) or a "value network." This is an opportunity for China's technological reform that has only been encountered in China for many years. In the past, China always complained about not using the Chinese Internet, because regardless of the network protocol or DNS, it is controlled by foreign countries. China's Internet market is large, but technological control is not in China.
Moreover, the blockchain has a concept of sovereignty, and the sovereign blockchain represents that Chinese laws can be automatically enforced on the Internet [8, 11, 12]. Such a chain network will be China's Internet, no longer a foreign Internet, but also a safe, reliable, high-speed Internet that can be used in financial markets, as well as public security laws and government affairs. Therefore, at the 2017 number fair, many speakers described the opportunity as “God Bless China”. At that meeting, the author also gave a speech on “Blockchain Internet” . Why do these people make such speeches and what is their opinion based on?
The traditional Internet can also do these applications, but it needs to add a lot of functions on the Internet, which is equivalent to establishing a virtual security environment in an insecure environment. This is not reliable. The chain network automatically provides these security encryption functions, and financial applications or important application development and use will be greatly facilitated. This is a blockchain Chinese dream, which is a good opportunity to build China's digital economy and digital society .
- Chip giant AMD joins blockchain game alliance to empower blockchain gaming
- 2019 financing report: DeFi replaces ICO, financing focus shifts to China
- Frequent trust crisis in the public welfare industry Can blockchain become a "life-saving straw"?
- Chinese universities will compile a batch of new teaching materials around blockchain, artificial intelligence, etc.
- Weekly: BTC plunges to drive down the market, blockchain project equity investment
- Ling listen to the micro-public bank blockchain open source observation: throw the code out, take the trust back
The Internet is not a foreign land, but now the Internet is not the Chinese Internet; but it is different after the chain, because this chain is the Chinese Internet. Don't think that this is an incredible thing, you can look at the British work in 2018. The British Law Commission (Legal Commission) began researching the use of blockchain and smart contract technology in the UK legal system in 2018. This will change the implementation and formulation of British law, and Chinese courts will not fall behind, and will begin to use blockchains in electronic evidence. These are huge changes, because this law and regulation appeared in code form, and automatically enforced Chinese law on the Internet, which is far from the current law formulation and implementation. This is also part of the blockchain Chinese dream .
If you suspect the necessity of the chain network, you can refer to the "2018 US version of the unified weights and measures: chain medicine supply chain" , the United States in 2012 because of the "fake medicine into the hospital" incident, causing many casualties, so in 2017 The research project was started in -2018, and finally a blockchain-based medical tracking project was proposed, and experiments began in 2019. There are 67,000 pharmacies, 8,000 hospitals, 3,000 approved drugs, and 200 drug manufacturers in the United States. If these are to be wound up, there will be a 4 x 10 14 combination. Such a huge system requires a chain network. In 2018, SWIFT conducted experiments, 34 banks participated (a total of 528 combinations), and the blockchain system used in the experiment could not be supported. If a blockchain can support 34 institutions, the US medical blockchain requires 75.7 billion chains (4 x 10 14 / 528 = 7.57 x 10 10 ). Isn't this chain combination a chain network?
1) Life is priceless, and the safety of medicine is priceless;
2) The US government FDA takes the lead in research and cooperates with related companies to propose a blockchain-based solution, which is equivalent to the US government and the pharmaceutical industry's identification (no currency) blockchain has practical value;
3) Goldman Sachs, the top investment company in the United States, took the lead in investing in this project, indicating that it has commercial value;
4) This is a currency-free blockchain project, which proves that the non-coin blockchain also has high commercial value;
5) This project also requires a lot of new technology development, and must lead the United States to develop new technologies; and this technology level is very high;
6) The US government's top-level system design is worth learning from other countries. On the one hand, it proposes that the US version of "Uniform Weights and Measures" unifies US pharmaceutical logistics management. On the one hand, it allows many units to protect people's lives under this system, and allows relevant participating companies to profit. , establish relevant ecology.
The chain will be full of chains, which will be the "value" of the Internet, unlike the previous "information" Internet. In the value network, it is important to be an asset, and it can also carry out a variety of asset transactions.
(1) The value network has the ability to be supervised, traceable, and authenticated, and can carry electronic evidence;
(2) Value network can support real-time low-latency transactions;
(3) The value network is scalable to support high throughput;
(4) Value Network allows transactions to be rollback. When the transaction is in error, the transaction can be rolled back within a certain period of time and within a certain range.
These characteristics are not based on the development of blockchains, but on the PFMI (Principles of Financial Market Infrastructures) [20, 21, 22]. This standard is supported by many central banks around the world, including the People's Bank of China, the European Central Bank, the Federal Reserve, the Bank of England, the Bank of Japan, the Bank of Canada and so on. This standard is the standard set by central banks in 2008 after the world economic crisis. An important indicator of PFMI is that when an economy experiences an economic crisis, the financial system can mechanismly prevent the crisis from spreading to other countries. PFMI is the principle of the universal financial system, not the principle designed specifically for blockchain systems.
PFMI was used by the Bank of Canada, the European Central Bank, and the Bank of Japan to evaluate the blockchain system in 2017-2018, and found that the current blockchain system cannot be evaluated by PFMI, including some features that have been considered to be blockchain strengths. Also did not pass. Traditionally, blockchain systems have claimed that blockchain systems are reliable because the same information is on multiple nodes. But the Bank of Canada believes that Ethereum and Corda (in fact, similar blockchain systems) are not reliable enough. This is very different from the public's idea of blockchain. The point is whether this problem can be solved?
The subsequent blockchain experiment from the Bank of Canada shows that these problems are not unsolvable, but that the blockchain system must be changed before it can be used by financial institutions. This is what the author has been saying in the past: the financial blockchain must be changed according to PFMI, not the PFMI. The current development trend of the financial blockchain is to transform the current blockchain system into a “financial blockchain” [21, 22].
Some of the blockchain systems now have PFMI principles, such as traceability, authentication, and electronic evidence. But overall, the current blockchain technology does not meet the following principles of PFMI:
According to the PFMI, all digital asset transactions are real-time and the regulation is real-time. Today, the public chain emphasizes the use of P2P to evade supervision, making it difficult for the public chain to conduct real-time transactions and supervision, which the government cannot support. The price of digital tokens fell sharply in 2018, also because the US SEC regulates digital tokens. If you take the P2P protocol, the blockchain can have real-time transactions and supervision, but there must be other mechanisms to support rollback. According to the needs of PFMI, the future financial blockchain will definitely change greatly, so the future blockchain and the current blockchain must be very different.
Because the chain network needs to support the financial blockchain, it must be fast, stable, reliable and protect privacy. Today, in stock trading, many units set up offices next to the exchange to reduce network information delays. They do this because the performance of the network protocol is now related to latency, the company is far from the exchange, and the information comes late. For example, when the company is out of the news, the stock will rise or fall very quickly. If there is a 0.5 second advantage, the price gap may be very large. And because the information conveyed is transaction information, the packet loss is very serious. In the blockchain system, if you drop a packet of data, there may be thousands of transactions lost at the same time, because a transaction fails, which will result in all the same blocks. The transactions failed, and they should have been successful. This may not only have problems with price differences, but may also lead to legal problems. These are the drawbacks of the current Internet protocol, which must be solved in the chain, and now there are solutions.
The chain network is only 4 years old today, but it also has different views and opinions:
There is also a view of the chain network. Some people think that the chain network is a network composed of DApp, blockchain system and distributed storage. If this is really a chain, this means that there are two Internets in the world, the chain (upper layer) + the traditional Internet (bottom layer), because on this model, DApp, blockchain system, distributed storage are all using traditional Internet protocols. In fact, it is the application of the Internet. According to this idea, any large-scale Internet application can become an upper-level Internet, so that the world will have many new Internet applications, such as Facebook, Alipay, WeChat, etc., which can claim to be the next-generation Internet because they are all large-scale Internet applications. . The author believes that the application is an application, unless the network protocol changes, and then the large application or application, not the new Internet.
To understand the design of the chain, you need to understand the history of the Internet. The following picture shows the Internet in 1971. You can see that there are fewer nodes. The participating units include Harvard University, MIT, Stanford University, UCLA, BBN, and RAND. This is a very small network. It was originally designed for the US military. Therefore, the network control protocol is designed to be black-boxed. It does not allow participants to control the internals of the network. These units can only control their own communication information outside.
Figure 1: Internet in 1971
Today's Internet has more than a billion nodes, and most of the Internet is now private, not military (the military network has its own private network). We can clearly see that today's Internet is very different from the Internet 40 years ago. This points to a serious question. Why is the Internet Protocol now similar to the agreement of the year? Is it that the design was perfect and adapted? The original black box agreement is still applicable today?
The following is the network architecture of the future chain network proposed by the author: the left picture is the current Internet protocol architecture, and the right picture is the chain network architecture. At the top of the left are some applications, such as content, e-commerce, video, social networking, etc.
Figure 2: Blockchain Internet architecture is different from traditional Internet
Figure 2: Blockchain Internet architecture is different from traditional Internet
In the future, there will be smart browsers and smart DApps that allow users to directly apply blockchain or the Internet of Things. The next layer was previously an interworking protocol between applications and applications. In the new Internet, there will be interworking protocols between blockchains; the next layer is the protocol of the blockchain itself, and the next layer is the network protocol.
From this picture, today's Internet is very different from the previous Internet. The Internet is now undergoing a change: the end-to-end design principles of the past are respected and the end-to-end principle is a good principle. But 12 years ago, Stanford University began to have different ideas and developed into a new network design trend, SDN (Software-Defined Networking) and NFV (Network Function Virtualization). SDN and NFV are different from the past Internet protocols. The biggest difference is that they no longer follow the end-to-end design principle, and that the Internet should not be a black box; because in the black box, collecting information and control systems is difficult. So there are obviously two schools in the network: the traditional school and the new school. There are also two schools in the chain.
The traditional school is represented by the Massachusetts Institute of Technology, who believes that traditional agreements on the chain do not need to be changed, such as the MIT Digital Tradecoin (tradecoin.mit.edu) digital currency project. They initially questioned the original Internet protocol design, but after the study, they thought that the end-to-end design was good and could continue to be used. They affirm three characteristics of the original Internet protocol: maintaining system survival, combining various types of services, combining various types of networks.
Readers may not know that the Internet end-to-end protocol was originally developed by the Massachusetts Institute of Technology's Computer Science & Artificial Intelligence Laboratory (CSAIL), so this MIT assessment is to evaluate colleagues' previous work. It was only this time from a different laboratory, the Media Laboratory, which also proposed the concept of “digital society”.
The MIT digital currency model is shown in the figure below. Each chain is an independent autonomous system. The left and right sides are homemade systems, and they are also blockchain systems.
Figure 3: MIT Digital Currency Architecture
The network structure still uses an end-to-end design (below): The right picture is an end-to-end design, each G is a Gateway representing a chain, and the left picture is an Overlay network. Basically they think of the chain as an end-to-end self-made blockchain system.
Figure 4: MIT uses traditional Internet protocols
Careful readers may find that this Tradecoin model is not the same as other digital currency models. For example, the stable currency model proposed by IBM in July 2018, the stable currency model proposed by JP Morgan Chase Bank in 2019, the Chinese Panda model, the Chinese golden monkey model, and the Bank of England's RSCoin digital currency model are not the same as the Tradecoin model. The model of IBM and JPMorgan Chase is centered on its own currency (digital dollar), and other currencies must exchange digital dollars; MIT, Panda model, and Golden Monkey model assume multiple independent digital currencies from the beginning, and provide various currency interactions. Mechanisms.
These models are not the same as the universe model and the Polkadot model. These two models start with digital tokens, but all use a centralized architecture to connect the participating chains. The MIT model, the panda model, and the golden monkey model are distributed models. A centralized chain, like a centralized pseudo-chain, is hard to accept in the future. In the early days, there was no choice because there was no better technology. But now there are some distributed chain architectures that will change. These centralized chains will change.
The other party did not accept the original idea and proposed the SDN/NFV concept because the Internet should not be a black box. They believe that a good Internet can be integrated control, not just between the end and the end: if you want to speed up the network and improve the overall network nature, you must do control within the network. This is exactly the opposite of the traditional network protocol.
They put forward new ideas and agreements. But from a certain point of view, they only propose old theories that already exist, such as circuit switching, virtual circuits, multi-commodity flow optimization, etc. they proposed. These were originally technologies of 30 to 40 years ago, and their commonality is integrated control.
Today, Google is promoting SDN/NFV because of its optimization of the data center network. They also got the same conclusions from Stanford University.
1. Today's agreement cannot provide timely control. It's easy to understand: all controls, because only end-to-end control, all control can only be done at the edge of the network, not in the network.
2. All protocols are only part of the control, not full control.
3. Because each protocol can only do a part, these protocols are becoming more and more complex.
These are all against PFMI. The financial system needs not only a control mechanism, but also a regulatory mechanism. The network is a complex system that does control the transmission of data packets, so control theory should be used.
1. Lack of "controllability": The mechanism on the network today has no way to directly or change the wrong behavior, because it is end-to-end.
2. Lack of "observation": Today only end-to-end observation, all information feedback is from the end to the other end, so that the information is not enough, and the information in the network can not be transmitted; the feedback information is transmitted to The other end is already delayed, far from the real-time target.
3. Lack of “structurability”: Because the entire network is treated as a black box, there is no way to make decisions in sufficient conflicting protocols and mechanisms.
In the history of Internet development, control theory experts did not participate. At first the Internet was designed by the telephone company and later by the computer company. Most of the experts in these two industries are not control experts. So that the Internet has huge room for improvement, for example:
1. The transmission of data packets on the Internet can be 10 to 1000 times faster than the current protocol.
2. Routing capacity can be increased by 50% to 80%.
3. In the system operation, the network can be monitored, controlled and managed by the whole system.
Among the Internet protocols, a special protocol is TCP (Transmission Control Protocol). TCP has a problem: its transmission volume drops due to the rise and change of RTT (round trip time).
Figure 5: TCP speed is greatly reduced due to delay 
This figure illustrates the amount of TCP transmission, which decreases exponentially as the RTT rises. Most people today think that this is a natural physical phenomenon. In fact, this is caused by poor protocol design, not a physical phenomenon. For example: If you buy 2 communication boxes, each box can transmit 10Gb/s. Connect the two boxes with a cable and you should get 10Gb/s. If the amount is not 10Gb/s, we say that this communication device is not good. Regardless of the cable delay of 1 millisecond, 10 milliseconds, 100 milliseconds or 1000 milliseconds, the box should maintain 10Gb/s, which is the specification of the communication device and should not be changed. But once TCP is added, it will not get 10Gb/s. This means that the normal amount of transmission should have nothing to do with the delay of propagation; the amount of transmission is degraded because of poor TCP design.
Another problem with TCP is that the amount of computation is too slow. Because TCP uses end-to-end iterations, the sender and the receiver need to send different data back and forth to enable the sender to calculate the optimal amount of transmission. If TCP requires N iterations, each time an RTT is needed (the time to propagate back and forth), the time it takes for the iteration to converge is N*RTT. But if the iteration of this calculation is placed within the network, for example, at or near the bottleneck link, information about congestion at the bottleneck can be easily obtained. Here, the calculation is done by iterations of the local computing device. The calculated optimal rate will be transmitted to 2 endpoints (sender and receiver). Because today's CPU speed is fast enough, the time to calculate the amount of transmission is negligible compared to the time it takes to send to the 2nd end. Therefore, the optimal throughput is only delayed by up to 1⁄2 RTT (which can be as short as 1⁄4RTT), which is significantly faster than N * RTT (where N is the number of iterations required for TCP to calculate its optimal throughput).
When Google discovered this, it developed a TCP alternative called BBR, which uses the deterministic equivalence control theory to improve TCP. Using this principle alone, it is 28 times better than the current TCP. This breaks the long-standing view of the online academic community that TCP is hard to make progress and that TCP cannot be replaced. Google’s contribution is to break this myth.
But the principle of deterministic equivalence is the old theory of 50 years ago in control theory. According to this principle: When there are many uncertainties in the system, take them as an average value, so that the uncertain system is regarded as a certain system, and the uncertain change is changed into the average value. The benefit is that because the system becomes deterministic, it is easy to design control; but the downside is that when the real system is not the same as the average, it can cause very bad control effects, which is the problem of BBR. The next figure shows that BBR performs poorly in a noisy path. This is a test done in the lab of the US Department of Energy. This test can see a lot of retransmissions of BBR than TCP: An example TCP retransmission number is 10700, but the number of BBR retransmissions is 240340 (24 times).
Figure 6: The number of retransmissions of BBR is too large 
If you use more mature control theory, you can increase the throughput of TCP by 10 to 1000 times, not just 28 times. Several companies today have been able to increase the speed of TCP by 10 to 1000 times. Signiant is a company that has developed a new protocol that transmits data 800 times faster than TCP. FileCatalyst is another company that has developed a new protocol that transmits data 1000 times faster than TCP (below). Here TCP is represented by FTP. The left side is 5 hours and 40 minutes, and the other is 1 minute and 23 seconds, which is more than 1000 times.
Figure 7: FileCatalyst increases the speed of TCP by a factor of 1000 
The other is called Aspera (buy by IBM). But Google's BBR is only 28 times. What is the difference here? IBM, FileCatalyst, and Signiant all need paths that cannot be shared with other traffic, but BBR can be shared with other traffic, but when BBR is shared with other traffic, it can't reach 1000 times, only 28 times.
I don't know if the reader observed an important phenomenon? That is, communication speed and delay have nothing to do with it. This means that the physical phenomenon that everyone thought was not a physical phenomenon, but the agreement was not designed.
Another important phenomenon is that traditional network protocol design is not suitable for the chain network, because TCP believes that dropping packets can be tolerated, and lost packets can be retransmitted. However, dropping packets in the new protocol is minimized. This is very important for the chain. In the consensus mechanism, if a packet is lost, the consensus mechanism may fail; if it fails, all transactions in the same block will fail together and must be re-started. This is different from the traditional system. In the traditional system, each transaction is handled separately. A transaction failure will only affect itself, but in the blockchain, any transaction in the block fails, and all transactions in the block fail together. Therefore, the data packets in the chain network cannot be lost at will. The faster the chain, the less it can lose packets, because each time it comes back, it can affect tens of thousands of other transactions.
The debate on network protocols today is not the beginning of the recent, nor the beginning of 20 years ago, it was started 40 years ago, and they are all in the same department. The theory of network usage control began with the MIT Laboratory for Information & Decision Systems (LIDS), which was developed by the Massachusetts Institute of Technology's Computer and Artificial Intelligence Laboratory. Both laboratories are in the same department, but in the same field there are two opposing schools.
The LIDS masters are gathered by world-renowned professors of control theory and information theory, R. Gallager, D. Bertsekas, S. Mitter, M. Athans, etc. Among them, Gallager and Bertsekas laid the theoretical foundation of modern networks. The second author is a student of two masters, and completed a doctoral thesis in this lab to study Internet protocols.
Figure 8: Robert Gallager, Professor of MIT LIDS
This school focuses on mathematics, modeling, and theoretical foundations, but rarely develops practical software. Due to the lack of actual software development, the theoretical impact of LIDS is small. And at the time the network was too small (64 nodes), the theory of control was not very useful, so it was not taken seriously. But when the 2nd author was at the MIT LIDS, the masters often mentioned the flaws in the network protocol, and thought that without the support of systemic mathematical theory, the Internet must be problematic in the future. After 40 years, the network is very large, and sure enough, the problem is that SDN/NFV aims to solve these problems. With the blockchain, the time for a major reform of the network protocol has arrived.
Although LIDS has no development protocol software, after more than 20 years of garage shop development, we have finally developed a network protocol and software based on control theory, and the actual performance is many times better than the current protocol. The original idea of LIDS today proved to be correct.
Many people think that TCP cannot be surpassed 10-1000 times. In fact, this is not only impossible, but also a natural phenomenon. Instead, an inefficient agreement is an unnatural phenomenon. It can be clearly seen from Figure 3 that 1000 times growth is very possible. When the delay is around 400ms, TCP performance drops to one-thousandth of the physical performance. If the performance and latency of the new network protocol is not related, the performance is 1000 times that of traditional TCP, and the world of such protocols has been made by several units, including our team.
In the past, the blockchain community has often seen three people become tigers, leading the academic and white clouds with commercial white papers. Under such circumstances, ordinary people will be misled. Sometimes even scholars do not find misunderstandings. When these misunderstandings are considered to be truths, even the scientific reports issued by government authorities will be teased in the media. For example, the British Chief Scientist Report is a famous example. It is a scientific report that strongly supports the blockchain, but it is ridiculed by the currency circle. It is believed that the most important scientific consultants in the British government do not understand the blockchain. These misleading factors affect the normal development of the blockchain and get unquestionable doubts and resistance.
Here are some misunderstandings:
1. Blockchain is decentralized and conflicts with national governance principles
Nakamoto's original text has no concept of "decentralization", and the decentralization of English text is not "decentralized". Most of the dictionaries are translated as "decentralized" rather than decentralized. “Decentralization” is the concept that later anarchists added. The Bank of Canada reported in 2017 that the blockchain system can be distributed or decentralized, but the center is still there, especially the central bank. There is no conflict between the blockchain and the current principles of government governance. The government is originally a decentralized operation. There are central and local governments, each with different powers. The use of the blockchain system does not mean that the center is removed, but the data consistency is assigned to the participating nodes to ensure that the participants and the center are synchronized. Chinese academics and research institutes should actively use the correct nouns to discuss blockchains.
2. Digital tokens are not blockchains, but an application of blockchains.
Some people negate the blockchain because they are disgusted with digital tokens, but they don't understand that digital tokens are not blockchains, just a blockchain application. The British Parliament has publicly declared that the British government will never use digital tokens such as Bitcoin in the government system. This declaration was expressed by the Bank of England, the chief scientist of the United Kingdom, the important think tank of the City of London, and related scholars after discussion in Congress. These are the units or scholars who support the blockchain in the UK. They all separate the blockchain from the digital tokens.
Instead, we should learn about the top-level institutional design of the United States. The US government is a country that strongly opposes digital tokens; in the early days, it even developed a digital token into prison. However, after the meeting of the Congress in early 2018, the US attitude was positive. Instead, the financial regulatory unit SEC, the CFTC issued a policy, and even the medical regulatory unit FDA issued policies and guidelines to support the blockchain. The US medical blockchain is advancing in this environment and building a lively ecosystem.
Finally, the chain will bring a huge dividend to China: the Internet has changed the Chinese society, bringing important technology and huge economic benefits to China. Chinese unicorn companies, including Huawei, Alibaba, Tencent, and Baidu, have brought significant productivity and economic benefits to China, all of which are related to the Internet. The chain network will be the next generation Internet. The new network protocol will not only bring high-performance network, but also make China have an autonomous network for the first time. This will be a sovereign network; the new Internet is not only a place of extra-legal, but can automatically enforce Chinese regulations and proactively maintain the rule of law in China.
.J. Ros & WK Tsai, “A theory of convergence order of maxmin rate allocation and an optimal protocol,” Proceedings IEEE INFOCOM 2001.
.J. Ros & WK Tsai, “A general theory of constrained max-min rate allocation for multicast networks,” Proceedings IEEE International Conference on Networks 2000 (ICON 2000).
.WKTsai, M. Iyer, & J. Ros, “Revisit the Strings Versus Clouds Debate for the Internet Architecture, Part II: QoS, Control, Management, and TCP,” Journal of Network and Systems Management, 2002 – Springer .
.Thomas Hardjono, Alexander Lipton, Alex "Sandy" Pentland, Towards a "Design Philosophy for Interoperable Blockchain Systems," MIT Connection Science, MIT, Cambridge, MA, USA, July 30, 2018
. Cai Weide et al., “One of the Chinese Dreams of Blockchain: Blockchain Internet leads China’s scientific and technological progress”,
. Cai Weide, Jiang Xiaofang, “The Chinese Dream of Blockchain: Building the Digital Society with the Chains of the World”,
. Cai Weide, He Juan, "2018 US version of "unified weights and measures" – chain network medicine supply chain management",
. Cai Weide et al., “The Blockchain Chinese Dream III: The automatic execution of law will subvert legal research, legal system and legal practice”,
. Cai Weide, Jiang Jiaying, “Three Important Principles of Smart Contracts”,
. Cai Weide et al., “Blockchain Internet”, 2017.6.3,
. Cai Weide et al., “Blockchain Internet Series (1): TCP end-to-end design is old and ill,”
. Cai Weide et al., “Blockchain Internet Series (2): Blockchain Internet Needs New Agreement”
. Cai Weide et al., “Blockchain Internet Series (3): Chain Network Needs Intranet Control and Monitoring”,
. Cai Weide et al., “Blockchain Internet Series (4): Chain network will be a complex network, and controlling time will be a key technology”
. Cai Weide et al., “Blockchain Internet Series (5): Chain network should consider adopting the path model”,
. Cai Weide, Kevin Tsai, “Blockchain Internet Series (6): One of the basic attributes of chain design: observability”,
. Cai Weide et al., “Application of distributed ledger technology in payment, clearing, and settlement services”,
. Cai Weide et al., “Application analysis of distributed ledger technology in bank payment, clearing and settlement business” Interpretation One of the “letter” PFMI series articles (1) from PFMI to see the blockchain,
. Cai Weide, Jiang Xiaofang, “Part 2 of the PFMI series: “The Road to Design” of the Clearing Chain”,