The problem of data security sharing and the problem of data islands is solved, and that is to use the blockchain.
More blockchain technology and application classification:
Blockchain applicationBlockchain development
Ethernet Square | Fabric | BCOS | cryptography | consensus algorithm | bitcoin | Other chain
Token Economy | Traditional Financial Scenarios | Decentralized Finance | Anti-counterfeiting Traceability | Data Sharing | Trusted Deposit
When sharing data between institutions in the same industry, there is a problem of "data islands". The most important thing is to ensure data security. Only on the basis of preventing copy leakage can it be possible to establish a mutual trust and sharing data platform. This is also an important and typical application of blockchain technology, that is, the current trend of combining AI + big data + blockchain. As the underlying platform, blockchain provides reliable protection for big data and data sources required by AI.
At present, the data of each company basically runs in its own domain. Each platform only has a part of its own information, rather than comprehensive information. The same is true between various industry organizations. They want to open up the connection between data and conduct data. Sharing is not easy. Because data has now become the core asset and wealth of the platform or enterprise, no one really wants to share it. Even if data is exchanged, some institutions will intentionally or unintentionally provide some low-quality data. When the data quality cannot be guaranteed, no matter how good the data model is, the correct result cannot be obtained. All in all, data sharing cannot be solved well under the Internet architecture, because data can be easily copied and transmitted, and data security cannot be guaranteed. Therefore, sharing data between platforms with data as their core asset is extremely risky.
The development of blockchain and encryption technology provides some new solutions to solve this problem. For example, the blockchain provides a credible "black box", so that everyone can put their data in without worrying about being copied or leaked. ; Or encrypt the data before sharing the data between the two parties, and then perform the calculation. The calculation result is also ciphertext. A method like this can not only complete the necessary data sharing and secure calculations, but will not leak the data or models to other people. To solve this problem technically effectively, the relevant data owners should have greater motivation for cooperation.
In the data sharing scenario, can a similar "data market" be formed among many "data sharing alliances", one is the providers of data and algorithm models, and the other is the buyers of data and algorithm models. The blockchain-based data market can effectively avoid data leakage and copying. Under the protection of the blockchain, data can be shared safely, and smart contracts in the blockchain are used to automatically execute some logic, such as undisputed statistics. Usage and pricing or data transactions.
Blockchain solution
Regarding data security sharing, there are several blockchain technology solutions:
① Homomorphic encryption
②Zero knowledge proof
③Multi-party secure computing (MPC)
④Trusted execution environment (TEE hardware)
Among them, ①-③ is cryptography technology, ④ is a technology combined with CPU hardware, and the more mature ones are homomorphic encryption and TEE technology.
For example:
Examples of cryptography solutions, a solution idea:
There are many participants in this data ecology. Each user needs to register on the blockchain platform and assign the corresponding public and private keys. The blockchain platform has the public key and digital certificate of each participant, so it can Easily verify the transactions of any authorized user, which is the key to data rights management. The original data is always private to itself, neither on the chain nor out of the domain. Only the address of the data is stored on the chain. The blockchain finds the data address through the requester's request, and requests the data from the data owner. Suppose that institutions A, B, and C each hold data, and A initiates a sharing request. When A needs B’s data for data sharing calculations, requester A sends a request transaction and attaches his own public key A. When the blockchain platform receives this transaction request, it verifies the requester, if not a legitimate user, rejects the transaction. If the verification is passed, the address of the data is queried through the smart contract. When the contract finds that the data address is B, request data from it. Data owner B finds the data by searching in its database. At this time, B uses his private key and requester A’s public key to generate a re-encryption key, which is used to encrypt the plaintext data retrieved from the database. Encrypted data is produced through encryption and returned to the blockchain. After the blockchain receives the encrypted data, it requests B's private key and the data that B participates in the calculation again, and at the same time opens the encrypted calculation sub-chain. This sub-chain is only used for secure calculations and is not known by any participants on the chain. High computing efficiency and less consensus verification ensure the efficient completion of computing tasks, and they will be automatically destroyed after completion. When the encrypted calculation sub-chain completes the calculation task, it returns the calculation result to the main chain. The chain then integrates the calculation results of other parties and returns the aggregate calculation result to A to complete the entire calculation task. In the entire blockchain system, smart contracts are used to count and evaluate the use of data. If a token incentive model is introduced, data contributors can automatically receive corresponding token rewards, forming a data market with free rules.
Original link: Thoughts on using blockchain for data security sharing