Scaling blockchain using asynchronous consensus zones
Technical field: A decentralized consensus system based on blockchain, in particular, a system capable of linear expansion in multiple independent and parallel Nakamoto consensus instances.
Background technology
- Blockchain technology provides a decentralized digital asset transfer infrastructure, but the challenges it faces include low transaction throughput (TPS) and the need for each node to repeat the communication, storage and state representation of the entire network, resulting in scalability and availability issues.
Content
- A new Nakamoto consensus system is proposed that can scale linearly without compromising decentralization and security.
- Scaling is achieved by running multiple independent and parallel Nakamoto consensus instances (called "zones"), with consensus in each zone being conducted independently, minimizing communication, sharing the workload of the entire network, and ensuring that each node is moderately burdened.
- Final atomicity is used to ensure the atomicity of cross-zone transactions, improving the efficiency of transaction completion without the overhead of the two-phase commit protocol.
- Batch mining is used to ensure that the effective mining capacity in each zone is consistent with the entire network, making the difficulty of attacking any single zone the same as attacking the entire network.
Example
- Supporting 48,000 nodes on a test bed containing 1,200 virtual machines, the technical effect shows a 1,000-fold increase in Bitcoin TPS.
Short answer questions
1. What is the main goal?
- Achieve linear expansion of the blockchain system by using multiple independent and parallel Nakamoto consensus instances while maintaining decentralization and security.
2. What are the key technologies proposed?
- Asynchronous consensus zone and batch mining technology to improve transaction throughput and efficiency by minimizing communication and final atomicity.
3. What existing blockchain technology problems are solved?
- Low transaction throughput (TPS) and the problem that each node needs to repeat the communication, storage and state representation of the entire network.
4. What are the experimental results?
- Supporting 48,000 nodes on a test bed containing 1,200 virtual machines, showing a 1,000-fold increase in Bitcoin TPS.