Wu Yanbing, Ph.D. of Tsinghua University, was inspired by Ulam’s “luck number” when he studied the energy consumption of “blockchain Impossible Triangle” and pow. It is found that using the characteristics of hash function can create a new consensus algorithm with ultra-low energy consumption, complete decentralization and high stability. On November 22, Wu Yanbing, founder of Ulam isomorphic chain, became the AMA of talk, a chain guest.
Let’s invite Dr. Wu to share the Ulam project with us
Ulam is a Polish mathematician, whose full name is Stanislav Marcin Ulam. He mainly invented an algorithm for prime number selection, and then gave our team a very good inspiration, and also a very good inspiration to tell us how to do this consensus.
In fact, the earliest time I found that uram put forward that prime selection is to select a number as a lucky number from a lot of numbers. Through certain methods, it is similar to the prime screening method. Then we can find that this is very similar to the mining in the blockchain. What kind of mining is it from a lot of numbers In many nodes, select one node at a time as the accounting node, and then get rewards to account for the whole network.
Now I will briefly introduce our Ulam algorithm and how it is implemented. We can see that if there are three nodes for bookkeeping, first we assume that the random number of the zero node is 10, the random number of the first node is 50, and the random number of the second node is 80.
Well, according to the different nodes, the lucky numbers of each node are also different. It can be obtained that they have different random numbers, just like the zero node, which has 10 lucky values, and it needs one random number, for example, its random number is 1, the first node, which has 50 lucky numbers, which is three random numbers, for example, 2, 5, 6, and the second node, which has 80 Fortunately, he has four random numbers, which are like 3, 4, 7 and 8.
Now, how can we decide the right to pack? For example, bitcoin uses POW to solve a problem. Whoever solves the problem first has the right to pack, which consumes computing power. But we are not the same. Without their energy consumption problem, we have a random number in the system, and then these three nodes, who owns the random number, will get the accounting right.
So let’s make a metaphor. If the random number 5 from the system this time, we can see that the first node has the random number 5 in it. Then the first node gets the right to pack and has the right to account. After the account is recorded, it will broadcast the block, and then the lucky value will be cleared.
This is a simple description. We can see that there will be several problems in this. For example, what should we do if the random numbers are the same or the random numbers out of blocks are not within the range of nodes? We use VRF random verifiable function and SM3 hash hash to ensure that the generated random number is not the same, and the random number generated by the system can fall within the range of random number owned by all nodes.
What can we guarantee at this point? First of all, the random numbers we have are different, which can ensure that our system will not generate bifurcations, because it is impossible for two nodes to reach the package at the same time, which can ensure that there will be no bifurcations. Another point is that the random number generated by the system can fall within the range of nodes, which can ensure that our miners will not generate empty packets, because the random number generated by the system will surely fall within the range of all nodes.
The advantages of our consensus algorithm come down to these three points
1、 The random numbers are different;
2、 The random number can fall within the range of all nodes;
3、 We use the improved VRF algorithm, which is random verifiable function. We can ensure that the generated random number is open and verifiable, without any tampering. Then anyone can verify it through the algorithm, and no one can make trouble. For example, when they open a lottery, they will carry out some operations, but this is completely impossible. Everyone’s random The numbers are verifiable, and the random numbers generated by the system are also verifiable.
Ulam’s six advantages:
1、 Super fragmentation;
2、 Complete decentralization;
3、 Never produce bitcoin like computing power centralization;
4、 49% of malicious nodes are fault tolerant;
5、 It’s also the most important. The performance is very high. TPS can achieve 1W +, transfer time is five seconds, and it’s basically impossible to generate bifurcations due to the ultra-low bifurcations rate. This is based on the fact that the generated random numbers I just mentioned are totally different;
6、 We have been proved by strict mathematics.
So let me explain how these six advantages are shown.
The first is the super fragmented node, which is based on our consensus. We can see our consensus. There is no computing power in it, so any device does not need to carry out hash competition, nor do it need to calculate the original image of hash, so any intelligent device can participate in it.
The second point is that we are completely decentralized, there is no centralized existence, and we can accommodate more than one million nodes, because all intelligent devices can participate.
Third, because we don’t have computational power, we won’t generate computational power centralization like bitcoin, because we are completely through the collision comparison of random numbers, and there is no mathematical problem to calculate the hash image.
The fourth point is that 49% of the malicious fault tolerance is the highest on the public chain.
The fifth point is that our TPS can reach 10000 +, now it’s 1W +, now our main network data is 1000 +, and then the confirmation time is five seconds. Why can we reach such a high TPS? When it comes to our original algorithm called non interactive transaction verification algorithm, we can guarantee that the TPS can reach a very high level, and the confirmation time can also reach a very short level.
Let me briefly introduce what non interactive transaction verification algorithm is. If a block is generated by bitcoin or Ethereum, there are 1000 transactions on the block to broadcast, and other miners need to verify these 1000 transactions. Ulam uses the method of knowledge proof to construct a non interactive transaction verification algorithm. To prove knowledge is simply to persuade the verifier to believe that he knows a certain “knowledge” (information). The general proof method is to construct polynomial time knowledge extractor to extract “knowledge”. Ordinary blockchain needs all miners to verify the transactions in the block when verifying the transactions. Ulam’s non interactive transaction verification algorithm generates a transaction verification certificate after miners pack the block and verify the transactions in the block. Other miners only need to verify the certificate generated by the packing node without verifying the transaction in the block, which greatly improves the TPS. In the original mode, if a block has 1000 transactions, all miners need to verify the 1000 transactions. This will undoubtedly take a lot of time to reduce TPS. Using Ulam’s original non interactive transaction verification algorithm, only one miner needs to verify 1000 transactions, and other miners only need to verify the certificates generated by the packing node. As a result, the speed will be 1000 times faster than before..
So far, it is limited to the network. If the national network facilities are well prepared, then our TPS will reach a higher level.
The last and most critical point is that our consensus and non interactive transaction verification algorithms are mathematically proven, and our proof has also been published and included in the paper.
Unlike other public chain projects, they just do a simple test, and then prove that they are safe, but we are through strict mathematical proof and various laboratory tests, and finally determine that it is safe.
This picture is our mathematical proof, in which we also published a paper. Then the first author of the paper is me, and my tutor also participated in the related work.
We can see that compared with traditional mining, if you are going to dig a bitcoin, what do you need now? You need a mine. There are hundreds or even thousands of mining machines in this mine. You can only dig a bitcoin. If you use a computer now, you can’t dig a bitcoin at all. You can’t dig a bitcoin for hundreds or thousands of years. But we uram can only participate in it through Internet connected smart devices, such as mobile phones, computers, and so on Ben is very, very small. You only need 2000 urams in it, and you will have the right to pack.
This is a process of our development. We have carried out some theoretical research in the laboratory since January last year. The consensus was published in June. In October, Tsinghua Chuang +, an internal investment institution of Tsinghua, only invested in Tsinghua team projects. After the first round of investment, Dade capital also invested in the first round. On January 13, we held the first press conference, and on the second day, we conducted the first internal test. Before the official launch, we had five internal tests, each time with different changes, basically a big change. Then we went online on April 8, joined the non interactive transaction verification algorithm on August 1, and in November Add our new function of synchronization chain.
This is our browser. You can see that the number of online mining nodes has reached more than 40000.
On the right, this is a case of our mining. As long as we keep networking on this page, we can mine.
This is a case of mining on our mobile phone.
This is our plan, the beginning, the development of our consensus. After some theoretical data and some proof tests in the laboratory, and then some papers are published, Tsinghua Chuang + investment was obtained in October, and it was formally started after the investment was obtained in December.
In January, we conducted the first internal test. At that time, the version was very simple. In February, we launched a public test. In April, we basically reached a consensus. On April 8, we officially launched the test. Then in August, we added the non interactive transaction verification algorithm to it, making the first TPS 462 up to 1017 now After that, it will gradually rise to 10000 yuan. Then in November, we developed this new function – isomorphic chain. Later, I will introduce what our isomorphic chain is. In 2020, that is, next year, we will develop smart contracts.
Now I can briefly introduce our new function – isomorphic chain. Why do we want to take the name of isomorphic chain? In fact, there is a concept of isomorphism in mathematics.
In mathematics, isomorphism can be basically understood as having the same properties and different forms of expression, so we use the name of isomorphic chain to define our strategy. They can also have their own mining attributes, or they can use our consensus of uram, TPS is close to 1W +, and the confirmation time reaches five seconds, but they can have their own different attributes, For example, in the aspect of big health, we should change it to fit big health, and then make some changes. In the aspect of sharing economy, we should make some changes.
We can see that the scenarios involved so far are big health, Internet of things, transportation, travel, games and sharing economy, but they are not limited to these aspects.
Each industry can design its own block time, total amount and minimum participation number of total rewards, which can be set, and they can upload their own data on the chain, such as medical data in the medical industry, such as equipment in the game industry, which are all OK.
We can see that this is an architecture of our isomorphic chain. First of all, we can see from the middle that this is a main chain of uram. We can use uram’s public chain to mine. For example, we can set up an Internet of things and a big healthy isomorphic chain on it. They can also use uram’s consensus to mine. For example, when we get the third block, we will make a hash function of it The hash value is then stored in our uram main chain, so that the uram main chain can ensure a security attribute of the isomorphic chain, and the problem of the isomorphic chain will not affect the main chain.
The isomorphic chain can have its own app, smart contract, etc. the isomorphic chain is equivalent to a public chain, but its security is guaranteed by uram. For example, after saying several blocks, it will pay a certain handling fee to the main chain of Ulam, which is also completely obtained by uram miners.
In this way, it is isomorphic chain, which is anchored to the main chain of uram through a one-way mapping.
We can see that this picture has all the same properties as Ulam’s, such as decentralization. He can also have his own independent nodes to mine, and can set any parameters according to his own needs, such as block time, block reward, and the minimum amount of participation. Then he is completely credible and tamper proof, such as being used in large scale In the field of health, we can upload some data of doctor’s operation to hash and upload it to the chain, so as to ensure that the original data will not be modified in any case, and we can judge some doctor-patient disputes through this, and one thing is distributed storage. All these data are completely existing in each mining section Order the top one. Then the data can also be linked and have a mining attribute, but the attribute can be selected or not. For example, the isomorphic chain can be used to mine or not to mine.
This is the page display of isomorphic chain
After clicking the details, we can also see some specific information on the isomorphic chain, such as the total amount of distribution and the address, precision and website of one of its isomorphic chains, which can be seen.
This interface is very similar to Ulam.
This is a brief introduction of mobile terminal isomorphic chain.
My introduction is over.
OK, thank you very much for the wonderful sharing that Dr. Wu has brought to us. I believe you all benefit a lot. I believe you need to consume so much. Thank you, Dr. Wu.