Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Optimizing View Change for Byzantine Fault Tolerance in Parallel Consensus

Yifei Xie, Btissam Er-Rahmadi, Xiao Chen, Tiejun Ma, Jane Hillston

TL;DR

This work addresses the slowdown caused by view changes in parallel BFT systems by formulating a View Change Optimization (VCO) as a mixed-integer program that jointly optimizes leader selection, follower assignment, and backup-leader provisioning across multiple committees. It introduces a decomposition-based solution with improved Benders cuts and an iterative backup-leader selection algorithm that reduces the view-change problem to efficient subproblems, including a 1-median subproblem for backups. The authors demonstrate substantial performance gains in Azure-based experiments under both normal and faulty conditions, showing improvements in throughput and latency as network size grows and failures occur. The approach offers a practical and scalable path to higher-performance parallel BFT systems with improved resilience to leader failures.

Abstract

The parallel Byzantine Fault Tolerant (BFT) protocol is viewed as a promising solution to address the consensus scalability issue of the permissioned blockchain. One of the main challenges in parallel BFT is the view change process that happens when the leader node fails, which can lead to performance bottlenecks. Existing parallel BFT protocols typically rely on passive view change mechanisms with blind leader rotation. Such approaches frequently select unavailable or slow nodes as leaders, resulting in degraded performance. To address these challenges, we propose a View Change Optimization (VCO) model based on mixed integer programming that optimizes leader selection and follower reassignment across parallel committees by considering communication delays and failure scenarios. We applied a decomposition method with efficient subproblems and improved benders cuts to solve the VCO model. Leveraging the results of improved decomposition solution method, we propose an efficient iterative backup leader selection algorithm as views proceed. By performing experiments in Microsoft Azure cloud environments, we demonstrate that the VCO-driven parallel BFT outperforms existing configuration methods under both normal operation and faulty condition. The results show that the VCO model is effective as network size increases, making it a suitable solution for high-performance parallel BFT systems.

Optimizing View Change for Byzantine Fault Tolerance in Parallel Consensus

TL;DR

This work addresses the slowdown caused by view changes in parallel BFT systems by formulating a View Change Optimization (VCO) as a mixed-integer program that jointly optimizes leader selection, follower assignment, and backup-leader provisioning across multiple committees. It introduces a decomposition-based solution with improved Benders cuts and an iterative backup-leader selection algorithm that reduces the view-change problem to efficient subproblems, including a 1-median subproblem for backups. The authors demonstrate substantial performance gains in Azure-based experiments under both normal and faulty conditions, showing improvements in throughput and latency as network size grows and failures occur. The approach offers a practical and scalable path to higher-performance parallel BFT systems with improved resilience to leader failures.

Abstract

The parallel Byzantine Fault Tolerant (BFT) protocol is viewed as a promising solution to address the consensus scalability issue of the permissioned blockchain. One of the main challenges in parallel BFT is the view change process that happens when the leader node fails, which can lead to performance bottlenecks. Existing parallel BFT protocols typically rely on passive view change mechanisms with blind leader rotation. Such approaches frequently select unavailable or slow nodes as leaders, resulting in degraded performance. To address these challenges, we propose a View Change Optimization (VCO) model based on mixed integer programming that optimizes leader selection and follower reassignment across parallel committees by considering communication delays and failure scenarios. We applied a decomposition method with efficient subproblems and improved benders cuts to solve the VCO model. Leveraging the results of improved decomposition solution method, we propose an efficient iterative backup leader selection algorithm as views proceed. By performing experiments in Microsoft Azure cloud environments, we demonstrate that the VCO-driven parallel BFT outperforms existing configuration methods under both normal operation and faulty condition. The results show that the VCO model is effective as network size increases, making it a suitable solution for high-performance parallel BFT systems.
Paper Structure (25 sections, 1 theorem, 16 equations, 5 figures, 1 table, 2 algorithms)

This paper contains 25 sections, 1 theorem, 16 equations, 5 figures, 1 table, 2 algorithms.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Leader-based BFT in parallel consensus
  4. Committee leadership and view change
  5. Reliable network optimization under failure
  6. Preliminaries
  7. Parallel Byzantine consensus network
  8. View change protocol
  9. Problem definition and assumptions
  10. View change optimization model
  11. Decision variables
  12. Objective function
  13. Optimization under normal case
  14. Optimization model for view change
  15. Backup node selection as views proceed
  16. ...and 10 more sections

Key Result

Proposition 5.1

For any binary $\hat{x}$ and any $i \geq 1$, the subproblem $PS(\hat{x}, S_i)$ has the integrality property.

Figures (5)

  • Figure 1: View change protocol phases under leader node failure.
  • Figure 2: View change protocol phases under leader node failure.
  • Figure 3: Performance comparison of parallel BFT algorithms
  • Figure 4: Performance comparison of ParBFT algorithms under failure condition
  • Figure 5: Performance comparison of ParBFT algorithms with various message sizes

Theorems & Definitions (1)

  • Proposition 5.1