FabricTL:Improving Blockchain Performance by Introducing Topological Sort with Evaluation of Transaction Influences
Liang Ye, Shaqi Fu, Wenhao Zhu, Wei Shao · 2025
The EOV blockchain architecture has been widely adopted across various industries, including finance, the Internet, and supply chains. EOV incorporates parallelism in the execution phase, which significantly enhances system throughput. However, under high concurrency conditions, its parallel execution may cause numerous MV CC conflicts during validation, leading to reduced throughput. To address the throughput limitations of EOV under high-concurrency settings, this paper proposes Fab-ricTL, which improves blockchain throughput by optimizing the sorting and validation phases. In the sorting phase, an improved topological sorting algorithm is employed. By prioritizing trans-actions based on node out-degree and in-degree, independently of others, it reduces conflicts and increases transaction success rates. In the validation phase, a circular buffer mechanism is used to retain parsed data, minimizing redundant parsing over-head. Experiments were conducted using standard blockchain performance testing tools. The results indicate that under high concurrency, FabricTL achieves throughput 1.12 times higher than Fabric and 1.05 times higher than Fabric++. Furthermore, FabricTL's sorting phase operates 50 % faster than Fabric++.