Grouped Multilayer Practical Byzantine Fault Tolerance Algorithm: A Practical Byzantine Fault Tolerance Consensus Algorithm Optimized for Digital Asset Trading Scenarios

Jian Liu, Wenlong Feng, Mengxing Huang, Siling Feng, Yu ZhangSensors 2023 · DOI

TL;DR

This paper introduces GM-PBFT, a novel consensus algorithm designed to improve the efficiency and scalability of blockchain-based digital asset trading. By grouping nodes and employing a hierarchical consensus structure combining PBFT and Raft, GM-PBFT significantly reduces communication overhead and transaction latency compared to traditional PBFT, while maintaining robust Byzantine fault tolerance.

What Problem Does This Paper Address?

The rapid growth of digital asset trading, especially within evolving virtual environments like the metaverse, presents significant challenges for existing blockchain consensus mechanisms. Traditional algorithms, such as Practical Byzantine Fault Tolerance (PBFT), struggle with excessive communication complexity and low consensus efficiency as the number of nodes increases. This leads to poor scalability, increased latency, and potential security vulnerabilities in large-scale, high-transaction-volume scenarios. The paper aims to address these limitations by proposing a more efficient and scalable consensus algorithm tailored for digital asset trading.

Methods

The core of this paper is the Grouped Multilayer Practical Byzantine Fault Tolerance (GM-PBFT) algorithm. It builds upon the PBFT algorithm by introducing several key innovations:

  • Node Grouping Strategy: Transactions nodes are categorized and grouped based on their attributes and experience. This allows for parallel processing of different transaction types within their respective groups, enhancing overall consensus efficiency. An ant colony optimization algorithm is used to identify optimal nodes within groups for supervisory roles.
  • Supervisory Strategy: To bolster Byzantine fault tolerance, a supervisory mechanism is implemented. Supervisory nodes, often identified through the ant colony algorithm, monitor the behavior of leader nodes within each group. This involves validation, auditing, and re-election processes, leveraging the Raft consensus algorithm to detect and mitigate malicious node attacks.
  • Grouped Multilayer Consensus: The consensus process is stratified into multiple layers. Upper layers utilize PBFT for consensus, recursively sending requests to lower layers until the final layer. The final layer employs the Raft consensus algorithm, with supervisory nodes ensuring the integrity of the leader's actions. This hierarchical approach breaks down complex consensus into manageable sub-problems, reducing communication complexity.

Datasets

Not discussed in the paper. The experiments were conducted using a simulated blockchain network.

Key Insights and Ideas

  • Hierarchical Consensus is Key: Breaking down a large-scale consensus problem into smaller, layered groups significantly reduces communication overhead. This is a fundamental shift from monolithic consensus approaches.
  • Raft and PBFT Synergy: Combining the Byzantine fault tolerance of PBFT with the efficiency and scalability of Raft (for the final layer) offers a powerful hybrid solution. Raft handles the operational consensus, while PBFT provides the foundational security.
  • Supervisory Nodes as a Security Layer: The introduction of dedicated supervisory nodes, managed by Raft, adds a crucial layer of defense against malicious actors, particularly in detecting and mitigating leader misbehavior.
  • Ant Colony Optimization for Node Selection: Using an ant colony algorithm to identify optimal supervisory nodes within groups is an interesting application of metaheuristics for improving network efficiency and trust.

Key Results

The experimental results demonstrate that GM-PBFT significantly outperforms the traditional PBFT algorithm in several key areas:

  • Communication Overhead: GM-PBFT shows a substantial reduction in communication overhead, especially as the number of nodes increases. For instance, with 30 nodes per layer, GM-PBFT's communication count is significantly lower than PBFT's.
  • Consensus Delay: The consensus delay is considerably reduced with GM-PBFT. While both algorithms' delays increase with more nodes, GM-PBFT's increase is more gradual, indicating better scalability.
  • Throughput: GM-PBFT achieves higher transaction throughput (TPS) compared to PBFT, making it more suitable for environments requiring high transaction processing capabilities.
  • Fault Tolerance: While PBFT has a fault tolerance rate of N/3, GM-PBFT, by incorporating Raft and supervisory nodes, offers improved fault tolerance, exceeding N/3.

Limitations and Future Work

The authors acknowledge that the technique is still in the experimental stage. Future research directions include:

  • Enhancing GM-PBFT Performance: Further optimization of the GM-PBFT algorithm to improve system consensus.
  • Addressing PBFT Node Limitations: Investigating solutions for the inherent limitations of PBFT regarding the number of consensus nodes.
  • Polycentric Multilayer Hierarchical Structure: Exploring the benefits of a more complex, polycentric multilayer hierarchical structure to further boost consensus efficiency in large-scale networks.
  • Application-Specific Solutions: Developing more tailored solutions based on the actual results of various blockchain applications combined with the enhanced PBFT algorithm.

Summary updated

Abstract

Based on the practical Byzantine fault tolerance algorithm (PBFT), a grouped multilayer PBFT consensus algorithm (GM-PBFT) is proposed to be applied to digital asset transactions in view of the problems with excessive communication complexity and low consensus efficiency found in the current consensus mechanism for digital asset transactions. Firstly, the transaction nodes are grouped by type, and each group can handle different types of consensus requests at the same time, which improves the consensus efficiency as well as the accuracy of digital asset transactions. Second, the group develops techniques like validation, auditing, and re-election to enhance Byzantine fault tolerance by thwarting malicious node attacks. This supervisory mechanism is implemented through the Raft consensus algorithm. Finally, the consensus is stratified for the nodes in the group, and the consensus nodes in the upper layer recursively send consensus requests to the lower layer until the consensus request reaches the end layer to ensure the consistency of the block ledger in the group. Based on the results of the experiment, the approach may significantly outperform the PBFT consensus algorithm when it comes to accuracy, efficiency, and preserving the security and reliability of transactions in large-scale network node digital transaction situations.
BibTeX
@inproceedings{liu2023grouped,
  author = {Jian Liu and Wenlong Feng and Mengxing Huang and Siling Feng and Yu Zhang},
  title = {{Grouped Multilayer Practical Byzantine Fault Tolerance Algorithm: A Practical Byzantine Fault Tolerance Consensus Algorithm Optimized for Digital Asset Trading Scenarios}},
  booktitle = {Sensors},
  year = {2023},
  doi = {10.3390/s23218903},
}
Figure