Parallel Blockchain Transaction Processing via Data Structure Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current blockchain transaction systems process transactions serially, limiting their performance and transaction per second (TPS) capacity, which is inadequate for handling larger volumes of transactions.
Innovation Solution
Implementing a method for parallel-processing blockchain transactions by grouping candidate transactions into independent groups, creating copies of the blockchain's data structure, executing these groups in parallel using a multi-core processor, and merging the updated copies to form a new block, thereby avoiding the need to lock the data structure during processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If transactions are processed serially with data structure locking, then data consistency is maintained, but transaction processing speed and TPS are limited
Solution Approach 1:
The patent segments the single data structure into multiple copies (first copy, second copy, etc.), allowing different transaction groups to process simultaneously on different copies. This segmentation enables parallel processing while maintaining data consistency through selective merging of updates.
Solution Approach 2:
The patent creates multiple copies of the blockchain data structure (state tree) before processing transactions. These copies serve as isolated working environments for parallel transaction execution, eliminating the need for locking the original data structure and enabling concurrent processing without compromising data integrity.
2Productivity
If multiple copies of data structure are created for parallel processing, then transaction processing efficiency is improved, but system resource consumption increases
Solution Approach 1:
Instead of creating complete independent copies of the entire blockchain state, the patent segments the data structure into relevant portions and creates copies only of those portions needed for parallel processing. This reduces memory overhead while still enabling concurrent transaction execution.
Solution Approach 2:
The patent applies different quality levels to different parts of the data structure - full copies are created only for the portions of the state tree that will be modified by transactions, while other portions are shared or reconstructed as needed. This optimizes resource usage by avoiding unnecessary duplication of unchanged data.
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for parallel-processing blockchain transactions are provided. One of the methods includes: obtaining a plurality of candidate transactions for adding to a blockchain; grouping the candidate transactions into one or more transaction groups; creating one or more copies of at least a portion of a data structure of a latest block of the blockchain; associating the one or more transaction groups respectively with the one or more copies of the data structure; executing the candidate transactions in each of the transaction groups and updating the associated copies of the data structure; and merging the updated copies of the data structure to obtain at least a portion of a new data structure of a new block to add to the blockchain.