Trust assurance method in blockchains
The main task for the blockchain in any information system is to ensure trust in an environment that excludes such trust by default. The problem of the “Byzantine generals” is often cited as an illustration of this problem. I will not dwell on it in detail, there are many materials on the Internet. In short, this task describes the decision-making process in an absolute anarchic environment, where everyone is for himself and believes in no one.
But it’s totally impossible not to trust anyone, sooner or later even the most convinced anarchist must decide on trust in another individual. In the case of cryptocurrency, as the first practical use of the blockchain technology, such trust is based on the agreement (consensus) of the overwhelming majority of users regarding the financial transactions that have occurred.
It should be clarified that consent is required not from ordinary users of cryptocurrency or, in general, information systems using blockchain technology, but among miners who specialize in forming next blocks of the chain and holders of a full copy of the blockchain. As the block chain grows, the relative number of subjects defining consensus decreases. The process of formation and adoption of new blocks from the equal model tends to be centralized. This is what potentially creates a situation where ordinary users' confidence in a particular cryptocurrency or information system can be greatly shaken.
At the initial stage of development of systems using blockchains, for example, Bitcoin, anyone could participate in the number of persons participating in the development of a consensus. The rule has not changed today, but the barrier threshold for full participation in this process has gradually grown. This is a noticeable increase in the size of the blockchain itself and an increase in computing power requirements.
')
The reality is that in the above-mentioned Bitcoin system, only a small number of players remain who form a consensus. According to the site Blockchain.info it is easy to verify this.
Current data on the distribution of computing power is available at blockchain.info/pools . As you can see, the real number of players does not exceed two dozen, and only 5-6 dominate. In the case of Bitcoin, more than 80% of computing resources are concentrated in one well-known Asian country.
This situation opens up a far from hypothetical possibility of manipulating the blockchain in the interests of a certain group of individuals. But not only Bitcoin, all current implementations of blockchain-based information systems to some extent suffer from this disadvantage. Existing and newly proposed methods of trust, such as PoW, PoS, DPoS, PoA, PoB, PoC, etc., are also not spared from such problems. Sometimes manipulations are “sanctified” by a large part of the users of the system, as was the case with the rollback in Ethereum to restore justice after the theft of coins on a large scale. This not only led to the fork by the disgruntled, but also questioned the use of Ethereum as a medium for smart contracts. Personally, I would not risk relying on this blockchain to confirm, for example, my rights to an apartment or land plot.
Despite the abundance of options for consensus within the blockchain, there is always the likelihood of voluntarism in deciding on the data stored in the blockchain. In blockchains based on the PoW consensus mechanism, which requires tremendous computing power and equally huge energy costs, the probability of correction is the smaller, the older the block. But the further, the PoW becomes an extremely costly and inefficient way to ensure trust. It protects the past from “erasures”, but under conditions of centralization, de facto does not guarantee the safety of the current.
Alternative methods designed to reduce the dependence on computing power, allow re-write the history of the blockchain in a relatively short time. After all, these methods artificially limit the number of participants in the consensus and the required computing power.
Do not forget that blockchains can be created and maintained without any consensus at all, for example, blockchains in independent corporate information systems or in state information systems. It is clear that such systems can generally do without the blockchain, whose role is precisely to guarantee the degree of trust. If a user trusts a state institution as a whole, then he can trust his information system. But on behalf of the state or a large company individual officials or employees act, so it cannot be excluded that, due to malicious intent or trivial errors, the information will be distorted. In the digital world, this can turn into a catastrophe for an individual. Therefore, the question of the degree of trust and guarantees of immutability of data become vital. Sooner or later, this will push government agencies and private companies to implement blockchains in their information systems.
A natural question arises: is it possible to ensure a guaranteed level of trust without using excessive computing power and in conditions of real centralization when creating a chain of blocks?
There is such a way; it consists in changing the principle of consensus. Currently, the blockchain consensus on a separate information system is an internal process.
And if you replace the internal consensus on the external? More precisely, not a consensus in the original sense, but more broadly, to delegate confirmation of trust in other information systems. In other words, separate and independent from each other blockchains voluntarily exchange the values ​​of the hashes of the currently formed blocks. Received hashes can be placed in blocks in the form of special records, special transactions. This does not require a radical restructuring of the blockchain-system operation algorithms, but will guarantee the trust on the principle of all-round bail . In the free Russian translation, this sounds like a "mutual responsibility . "
Such a voluntary exchange of hashes solves a number of problems.
Firstly, the participation of any information system in this exchange makes it clear to users that they receive guarantees that the information previously entered in the blockchain of this system is unchanged. After all, thanks to the mutual exchange of hashes and the openness of the hashes themselves, it is easy to check whether there have been “backdating” any adjustments to the database entries. This is especially important for projects that do not involve the competition of miners. For example, for corporate and state blockchains. The fact of the participation of any blockchain in such a system of digital “circular guarantee” makes it attractive to the user.
Secondly, the need for an insane race of computing power disappears, since the guarantee of preservation of the main branch of the block chain is transferred to external systems. At the same time, the possibility of collusion between participants of the voluntary-mutual exchange of hashes is almost completely eliminated. The more of them, the less likelihood of a backstage agreement. The situation returns to the roots at a new qualitative level, instead of equality between users within the system, as the basis of consensus, equality between systems arises.
Third, it is possible to create blockchain systems of various topologies, not limited to the current “flat” P2P. Thanks to the delegation of trust, it is possible to create “multidimensional” structures, for example, hierarchically centralized systems with a single center of formation of blocks. This will solve the problem of low transaction speed inherent in, say, Bitcoin. The availability of various topological schemes will allow the introduction of blockchain technology into existing corporate, public and state information systems without their fundamental restructuring.
You can also enumerate the prospects opening up when introducing the mechanism of mutual exchange of hashes between independent blockchains, but I would especially note the possibility of transition to a new quality of smart contracts.
Close analogues of the proposed method is a system with permissioned blockchain (affiliated, dependent blockchains) in Ethereum, as well as the Exonum of Bitfury Group.
The mentioned decision on the mutual exchange of hashes between different chains of blockchains was developed as part of the Business Trust Rating System project. The project has not yet been completed, but a patent application has been filed for the key solution technology described above. Simultaneously with the mentioned project, work is underway on a non-commercial project that allows for the asynchronous exchange of hashes between independent blockchains and create a mechanism for verifying confidence in any blockchain by anyone.
In principle, independent blockchains can exchange hashes of current blocks directly, as indicated in the figure:
Such a scheme is quite possible, but it has several limitations.
First, low scalability. With an increase in the number of participants, many organizational and technical problems will arise in their interaction.
Secondly, even with two participants, there is a synchronization problem, because the period of the formation of blocks in each blockchain is its own, sometimes different in order.
Thirdly, the process of control over the accuracy of data by the user is not quite obvious.
An independent exchange center (DLT Trust) in the form of a non-profit and public organization is capable of leveling these problems. The principle of its operation is quite simple. Any blockchain A, after forming the next block, sends its hash to this exchange center and in response receives the current hashes of other blockchains (B and C) for the current moment. The blockchain writes the resulting hashes into the next block as a special transaction. Thus, distributed storage of hashes is achieved, duplicated by the number of exchange participants.
Currently, two storage options are being considered: a separate entry for each blockchain or a recording of the resulting hash, obtained using the Merkle tree algorithm.
The mechanism for verifying the validity of the hash of any block of any blockchain is transparent to the user who wants to make sure that the data in a particular block is immutable. To do this, it sends a request to the DLT Trust system, containing the attributes of the block of interest. The system for the specified attributes is looking for confirmation of the hash value of the selected block in the corresponding blocks of other blockchains. If everything matches, then the user receives a message about the immutability of previously recorded data.
This mechanism is able to reveal the fact of data substitution, for example, in blockchains formed without internal consensus or in conditions of limited internal consensus. The owner of such a blockchain has the ability to change the record of interest and rebuild the blockchain, starting with the modified block to preserve the integrity of the chain, but with new values. The presence of copies of the hashes of the previous version of the blockchain in a number of independent other blockchains makes this operation meaningless. One fact of a discrepancy is enough for users to lose their trust in such an information system forever.
UPD. The algorithm of the All-Round Bail method presented in this article can be tested using a demo example.
Consensus as the Achilles heel of the blockchain
But it’s totally impossible not to trust anyone, sooner or later even the most convinced anarchist must decide on trust in another individual. In the case of cryptocurrency, as the first practical use of the blockchain technology, such trust is based on the agreement (consensus) of the overwhelming majority of users regarding the financial transactions that have occurred.
It should be clarified that consent is required not from ordinary users of cryptocurrency or, in general, information systems using blockchain technology, but among miners who specialize in forming next blocks of the chain and holders of a full copy of the blockchain. As the block chain grows, the relative number of subjects defining consensus decreases. The process of formation and adoption of new blocks from the equal model tends to be centralized. This is what potentially creates a situation where ordinary users' confidence in a particular cryptocurrency or information system can be greatly shaken.
At the initial stage of development of systems using blockchains, for example, Bitcoin, anyone could participate in the number of persons participating in the development of a consensus. The rule has not changed today, but the barrier threshold for full participation in this process has gradually grown. This is a noticeable increase in the size of the blockchain itself and an increase in computing power requirements.
')
The reality is that in the above-mentioned Bitcoin system, only a small number of players remain who form a consensus. According to the site Blockchain.info it is easy to verify this.
Current data on the distribution of computing power is available at blockchain.info/pools . As you can see, the real number of players does not exceed two dozen, and only 5-6 dominate. In the case of Bitcoin, more than 80% of computing resources are concentrated in one well-known Asian country.
This situation opens up a far from hypothetical possibility of manipulating the blockchain in the interests of a certain group of individuals. But not only Bitcoin, all current implementations of blockchain-based information systems to some extent suffer from this disadvantage. Existing and newly proposed methods of trust, such as PoW, PoS, DPoS, PoA, PoB, PoC, etc., are also not spared from such problems. Sometimes manipulations are “sanctified” by a large part of the users of the system, as was the case with the rollback in Ethereum to restore justice after the theft of coins on a large scale. This not only led to the fork by the disgruntled, but also questioned the use of Ethereum as a medium for smart contracts. Personally, I would not risk relying on this blockchain to confirm, for example, my rights to an apartment or land plot.
The risks of data substitution
Despite the abundance of options for consensus within the blockchain, there is always the likelihood of voluntarism in deciding on the data stored in the blockchain. In blockchains based on the PoW consensus mechanism, which requires tremendous computing power and equally huge energy costs, the probability of correction is the smaller, the older the block. But the further, the PoW becomes an extremely costly and inefficient way to ensure trust. It protects the past from “erasures”, but under conditions of centralization, de facto does not guarantee the safety of the current.
Alternative methods designed to reduce the dependence on computing power, allow re-write the history of the blockchain in a relatively short time. After all, these methods artificially limit the number of participants in the consensus and the required computing power.
Do not forget that blockchains can be created and maintained without any consensus at all, for example, blockchains in independent corporate information systems or in state information systems. It is clear that such systems can generally do without the blockchain, whose role is precisely to guarantee the degree of trust. If a user trusts a state institution as a whole, then he can trust his information system. But on behalf of the state or a large company individual officials or employees act, so it cannot be excluded that, due to malicious intent or trivial errors, the information will be distorted. In the digital world, this can turn into a catastrophe for an individual. Therefore, the question of the degree of trust and guarantees of immutability of data become vital. Sooner or later, this will push government agencies and private companies to implement blockchains in their information systems.
A natural question arises: is it possible to ensure a guaranteed level of trust without using excessive computing power and in conditions of real centralization when creating a chain of blocks?
Bail as a kind of consensus
There is such a way; it consists in changing the principle of consensus. Currently, the blockchain consensus on a separate information system is an internal process.
And if you replace the internal consensus on the external? More precisely, not a consensus in the original sense, but more broadly, to delegate confirmation of trust in other information systems. In other words, separate and independent from each other blockchains voluntarily exchange the values ​​of the hashes of the currently formed blocks. Received hashes can be placed in blocks in the form of special records, special transactions. This does not require a radical restructuring of the blockchain-system operation algorithms, but will guarantee the trust on the principle of all-round bail . In the free Russian translation, this sounds like a "mutual responsibility . "
Such a voluntary exchange of hashes solves a number of problems.
Firstly, the participation of any information system in this exchange makes it clear to users that they receive guarantees that the information previously entered in the blockchain of this system is unchanged. After all, thanks to the mutual exchange of hashes and the openness of the hashes themselves, it is easy to check whether there have been “backdating” any adjustments to the database entries. This is especially important for projects that do not involve the competition of miners. For example, for corporate and state blockchains. The fact of the participation of any blockchain in such a system of digital “circular guarantee” makes it attractive to the user.
Secondly, the need for an insane race of computing power disappears, since the guarantee of preservation of the main branch of the block chain is transferred to external systems. At the same time, the possibility of collusion between participants of the voluntary-mutual exchange of hashes is almost completely eliminated. The more of them, the less likelihood of a backstage agreement. The situation returns to the roots at a new qualitative level, instead of equality between users within the system, as the basis of consensus, equality between systems arises.
Third, it is possible to create blockchain systems of various topologies, not limited to the current “flat” P2P. Thanks to the delegation of trust, it is possible to create “multidimensional” structures, for example, hierarchically centralized systems with a single center of formation of blocks. This will solve the problem of low transaction speed inherent in, say, Bitcoin. The availability of various topological schemes will allow the introduction of blockchain technology into existing corporate, public and state information systems without their fundamental restructuring.
You can also enumerate the prospects opening up when introducing the mechanism of mutual exchange of hashes between independent blockchains, but I would especially note the possibility of transition to a new quality of smart contracts.
Close analogues of the proposed method is a system with permissioned blockchain (affiliated, dependent blockchains) in Ethereum, as well as the Exonum of Bitfury Group.
An example of the decision on the principle of "circular bail"
The mentioned decision on the mutual exchange of hashes between different chains of blockchains was developed as part of the Business Trust Rating System project. The project has not yet been completed, but a patent application has been filed for the key solution technology described above. Simultaneously with the mentioned project, work is underway on a non-commercial project that allows for the asynchronous exchange of hashes between independent blockchains and create a mechanism for verifying confidence in any blockchain by anyone.
In principle, independent blockchains can exchange hashes of current blocks directly, as indicated in the figure:
Such a scheme is quite possible, but it has several limitations.
First, low scalability. With an increase in the number of participants, many organizational and technical problems will arise in their interaction.
Secondly, even with two participants, there is a synchronization problem, because the period of the formation of blocks in each blockchain is its own, sometimes different in order.
Thirdly, the process of control over the accuracy of data by the user is not quite obvious.
An independent exchange center (DLT Trust) in the form of a non-profit and public organization is capable of leveling these problems. The principle of its operation is quite simple. Any blockchain A, after forming the next block, sends its hash to this exchange center and in response receives the current hashes of other blockchains (B and C) for the current moment. The blockchain writes the resulting hashes into the next block as a special transaction. Thus, distributed storage of hashes is achieved, duplicated by the number of exchange participants.
Currently, two storage options are being considered: a separate entry for each blockchain or a recording of the resulting hash, obtained using the Merkle tree algorithm.
The mechanism for verifying the validity of the hash of any block of any blockchain is transparent to the user who wants to make sure that the data in a particular block is immutable. To do this, it sends a request to the DLT Trust system, containing the attributes of the block of interest. The system for the specified attributes is looking for confirmation of the hash value of the selected block in the corresponding blocks of other blockchains. If everything matches, then the user receives a message about the immutability of previously recorded data.
This mechanism is able to reveal the fact of data substitution, for example, in blockchains formed without internal consensus or in conditions of limited internal consensus. The owner of such a blockchain has the ability to change the record of interest and rebuild the blockchain, starting with the modified block to preserve the integrity of the chain, but with new values. The presence of copies of the hashes of the previous version of the blockchain in a number of independent other blockchains makes this operation meaningless. One fact of a discrepancy is enough for users to lose their trust in such an information system forever.
UPD. The algorithm of the All-Round Bail method presented in this article can be tested using a demo example.
Source: https://habr.com/ru/post/338696/
All Articles