PlatON Privacy-Preserving Computation White Paper | Part 1. Overall Framework
Since the PlatON white paper was released on 19 July 2018, and after nearly three years of continuous upgrades and improvements, PlatON has has grown into a distributed computing network infrastructure with complete architecture, leading features and excellent performance.
In the PlatON network, privacy-preserving computation technology is the core engine supporting PlatON and Alaya networks, laying the cryptographic foundation of the PlatON privacy-preserving computation AI network.
LatticeX Foundation aims to solve three main problems in the blockchain community, including privacy, storage and key management.
The LatticeX Foundation has presented ideas describing the underlying cryptographic components and potential capabilities to support privacy-preserving applications for both networks.
This article will be presented in several parts, today this part is about the algorithm, On-Chain Capabilities, and Off-Chain Services.
Part 1. Overall Framework
Blockchain community has grown very rapidly these years, and plenty of applications are re-built in decentralized systems. However, many urgent problems still exist and shall be solved in blockchain systems. This article aims to consider privacy, storage and key management, which are widely discussed in the community.
Dating back to the design of Bitcoin, all participants maintain the network by storing copies of all the data. Although it provides a potential method to achieve “consensus”, the transaction data is public to all the participants. Ethereum introduces smart contract that extends payment transactions into any computable functions. Developers could design their business logic using smart contract, once the condition is satisfied, the smart contract will be run automatically. Automatic execution essentially means that each node of the network executes the smart contract repeatedly in local. Therefore, the input, output and description of the function should be public to all participants.
Developers now realize privacy is becoming one of the most important issues when building financial business among enterprises with blockchains. The privacy of the transfer amount, the anonymity of payer and payee, and the privacy of the business logic (i.e., the function described in smart contract) should be protected in most scenarios. LatticeX Foundation aims to provide a privacy-preserving infrastructure for decentralized business among financial institutes.
As the name of blockchain, each block containing multiple transactions is chained to another with a cryptographic hash function. The one-wayness of the hash function and append-only structure of blockchain make it computationally hard to tamper data stored. However, as transactions continue to grow, so does the amount of data that needs to be stored on the blockchain. At the time of writing this manuscript, Bitcoin and Ethereum store hundreds of gigabyte amount of data, and the size is still increasing. It takes days on a typical laptop as a full node to synchronize and verify these historical transactions. The underlying consensus protocol of PlatON/Alaya is based on proof of stake (PoS). In a PoS system, the validators often run full nodes in a cloud provided by a third party, and almost over 50% cost of cloud services comes from storage. With the increasing of blockchain data, the cost of validator arises. Therefore, reducing the size of storage is important for validators to reduce cost and further maintain the security of the network.
LatticeX Foundation proposes a framework, as shown in Figure 1, to deal with privacy and storage issues in PlatON/Alaya. This solution or plan is based on the architecture of PlatON 2.0. The vision of PlatON 2.0 is to build a decentralized and collaborative AI network and global brain to drive the democratization of artificial general intelligence.
A three-layer network is considered in PlatON 2.0. The first layer is Consensus Layer. Consensus network is a decentralized network composed of blockchain nodes, which are connected to each other through P2P protocols and can be consensual through consensus protocols in an environment where no one needs to be trusted. On the blockchain network, smart contracts can be executed automatically. The second layer is the Privacy-Preserving Computation Network. Data nodes and computing nodes could connect to this layer and publish data and contribute computing power. Through smart contracts on the blockchain, a decentralized sharing and trading marketplace for data, algorithms and computing power can be built. Based on the cryptographic economics it will form an effective incentive mechanism to motivate more data, algorithms and computing power to join the network. The third layer is Collaborative AI Network, in which multi-agent systems and AI agents can operate independently, and finally form autonomous AI networks.
One could take the solution of this article as advanced properties of PlatON 2.0 in layer 1 and layer 2. It is designed dedicatedly to solve privacy and storage problems mentioned above.
Four levels are considered in this framework. The first level contains necessary cryptographic algorithms, especially related to zero-knowledge proofs. The second level is built on layer 1 in PlatON 2.0, it focuses on providing on-chain capabilities to solve the two problems above. The introduction of cryptographic tools will reduce the entire performance inevitably, for which the raw transactions need to be stored in some place anyway, instead of on the blockchain. The third level, which is built on layer 2, aims to provide computation and storage services off the chain, and thus we call it off-chain services. The fourth layer is built for privacy-preserving applications. Standard APIs, templates and protocols are created for different applications. More details of each level will be described in the coming sections.
In addition, one common requirement for all applications is to manage the private key securely. This article also provides a relatively independent solution for key management service in follow-up article.