| With the rise of the concept of blockchain in recent years,more and more people begin to engage in the crypto industry.The entire crypto industry is devoted into figuring out how to open the door to a decentralized world,allowing people to securely trade without relying on third parties.The so-called technology is also collectively referred to as distributed ledger technology.A distributed ledger is similar to a synchronous database that records the transfer of value between network participants.The significance of distribution is that third-party arbitration institutions are eliminated for this value transfer.The entire ledger records the history of valid transactions and cannot be tampered with.As early as 2008,Satoshi Nakamoto first published the Bitcoin white paper,which caught the attention of the world in the subsequent events.By the end of 2017,the price of Bitcoin once exceeded 100,000 RMB.But with the experience gained from practice and new cognitive updates,the use of bitcoin alone is far from satisfying people's ideal trading scenarios.The problems of the Bitcoin system itself also limit its application scenarios and development,and the mechanism itself determines its scope of use.Low throughput,high latency,and its inherent workload certification mechanism and high transaction fees make it impossible to meet small payments,just as no one wants to go to Starbucks to buy a cup of coffee for an hour to confirm that the transaction is successful and then have to pay Bitcoin twice the price of coffee in RMB.In addition,its own workload proof mechanism has also led to a large amount of energy consumption.Bitcoin's annual mining power consumption has exceeded the national electricity consumption in Iraq for one year.The occurrence of these problems has made people think about new systems and new mechanisms.In the more than ten years since 2008,a large number of programs have been introduced.Not only that,but many scholars and engineers have also released new consensus plans.For example,in 2012,Sunny King first proposed a proof-of-stake mechanism and applied it to the Peercoin.The proof-of-stake can effectively solve the problem of energy consumption and increase the speed of the block-issuing process.Later,there was a delegated-proof-of-stake algorithm,which greatly improved the speed of the system.For example,the subsequent EOS project adopted the delegated-proof-of-stake mechanism.However,all kinds of mechanisms are based on block chain,and no one seems to think about the problem of the chain-structure itself.In the inherent block chain structure,due to the existence of the block concept,transactions must undergo a process of package screening.Therefore,no matter which consensus algorithm is based,this part will inevitably produce partial consumption.Not only that,but the result of packing blocks will also cause some waste.For example,only the longest chain in the bitcoin is retained,and all the remaining branches will be ignored.And we know that the process of packing the block is a difficult process,which is labor or financial or energy.Therefore,if improvements are made under this structure,the inherent cost of the block cannot be avoided.Therefore,thinking about its own chain structure becomes meaningful.Until 2015,Sergio Demian Lerner proposed a new chain structure: a directed acyclic graph(DAG).This underlying structure breaks the inherent thinking of the original block,allowing each transaction to be immediately chained,and the entire network can be constrained by a new consensus algorithm.The topology of the entire network is more complex than the blockchain structure,which also makes the consensus algorithm more critical.A number of distributed ledger techniques based on directed acyclic graph structures,such as the IOTA released in 2016,and the subsequent Byteball,followed.Currently,for distributed ledger technology based on directed acyclic graph structure,the two are in a leading position.However,both of them have shortcomings in practice and theory.Among them,IOTA has a tendency to be centralized on the issue of final consistency.he agreement uses a manager similar to dictator to determine whether the transaction is consistent.In the Byteball project,both the double-spend problem and the final consistency issue are controversial.For example,the protocol talks about the choice of the main chain.Each main chain is relative to a certain node.From the different nodes back to the initial point,different main chains will appear.Therefore,in the white paper given by Byteball,the only main chain recognized by the whole network may not exist,and many main chains may not be unified in the end.Therefore,solving the double-spend problem through the order defined by the main chain may not work.In addition,for the final consistency problem,if it is not completely determined that the main chain will be normalized,it cannot be proved that all the main chains relative to the leaf nodes will eventually pass through.For the structure itself,the scheme given by Byteball may have a good effect on the specific implementation,but it is not rigorous from a mathematical point of view.In addition,from an economic point of view,the only main chain will also bring a lot of waste of resources.Events that are not in the single main chain will always be ignored,which is much the same as the waste caused by Bitcoin,even if there is no block packaging process.Similarly,from the perspective of specific implementation,the main network that it has established in the past two years has almost developed into a single chain.The results of this implementation also confirm that the agreement itself has certain problems,and it also shows that efficiency has not been achieved.Also,there are many uncontrollable factors,all of which depend on the imperfections of the agreement itself.However,it is worth mentioning that the witness mechanism and the choice of the backbone provide a good idea and vision for the distributed ledger technology based on the directed acyclic graph structure.Therefore,for this new underlying structure,although the related projects are emerging one after another,the consensus mechanism is still the focus of the discussion.Until now,there has been no practice to prove that there is a perfect solution based on the directed acyclic graph structure to efficiently meet the needs of real-world applications and to be as safe and decentralized as possible.In this article,we will focus on the consensus issue based on the directed acyclic graph chain structure.This paper will provide a new consensus solution to solve the double-spend problem and the final consistency(finality)problem.Some necessary definitions will be given at the beginning,and a detailed description of the distributed ledger will follow.The core and key points are how to effectively solve the double-spend problem and the final consistency problem,at the same time as safe,efficient and decentralized as possible,so that the whole network node can get more participation.Therefore,in the double-spend problem,a number of reputable proxy nodes are defined as endorsements of transactions.These credibility nodes generally act as credible characters in real life,similar to witnesses in Byteball,but the way to solve double-spend is different.Due to the structure itself,the transaction-issuing process will become very fast and easy,so the complexity should be reduced as much as possible in the double-spend problem,and the bad transaction between the two should be quickly eliminated..For the final consistency issue,since it is uniquely determined and cannot be tampered with,it should be treated with emphasis and as many nodes as possible involved.At the same time,it is also necessary to reach a consensus on the whole network.When most of the credit nodes and non-reputation nodes recognize it,the transaction will assume part of the entire network,so it is fully agreed.Different from Byteball,there is no unique main chain in the consensus mechanism introduced in this paper.This unique main chain is difficult to achieve due to its structural limitations.If there is a unique main chain,the final effective structure of the whole network.Will be tantamount to bitcoin or other primitive blockchain structure,which loses the meaning of the directed acyclic graph structure itself.Therefore,it is easy to achieve a unique main chain is a blockchain structure,such as bitcoin always chooses the longest chain as the main chain,and the packing blocks on it and the deliberately manufactured uplink delay are also for its structural service.In the directed acyclic graph structure,the goal is to achieve high efficiency and low cost while ensuring security and decentralization.Therefore,it is not necessary to find a unique main chain to add complexity to the entire network,and only need to reach a structural consensus.However,compared with the final consistency mechanism of Bitcoin,the scheme given in this paper can reach 100 percent consensus while in the mechanism given by Bitcoin,the probability that any transaction is confirmed by the final consistency will not reach1.In addition,in all distributed ledgers,according to the blockchain ”impossible triangle”,the three points of scalability,security and decentralization cannot be achieved at the same time.Therefore,you can only sacrifice one of them to save the remaining two as much as possible.The idea of ?using the agent ”eyes” is equivalent to sacrificing partial decentralization,resulting in relatively weak inequalities between nodes,but greatly guaranteeing security and scalability.With the structure of the directed acyclic graph itself,it is hoped that the performance will be relatively large.At the same time,under the consensus mechanism of the text,almost all honest and effective transactions issued by effective nodes can be recognized,and it is difficult to have such a large number of neglected situations in Bitcoin.After the introduction of the consensus part,the paper gives some incentives for the node,as well as the function and release of the token.This also plays a decisive role in the entire system.Just like the real world,the economy maintains the stability of the entire industrial society.In this distributed system,the economy acts as a way to maintain system balance.Nodes that receive incentives will be more willing to serve the system,and penalties for bad nodes will also eliminate some of the system's adverse events.At the end of this article,we discuss some possible attacks,such as agent attacks,and consider malicious nodes to buy proxy reputation nodes.This situation is unlikely to happen,but it must rely on the system's own constraints to eliminate it.For dynamic systems,the imaginary attack methods often do not occur.History tells us that attack methods are often not thought of by system builders.The main contribution of this paper is to provide a feasible consensus mechanism reference scheme for distributed ledger technology based on directed acyclic graphs.The program incorporates the advantages of previous projects such as Byteball,EOS Bitcoin,etc.,and effectively avoids the shortcomings known in previous scenarios.In the specific implementation process,it is only necessary to arbitrarily transform the parameters according to the method in the text. |
PDF Full Text Request