Two-Fold Byzantine Fault Tolerance Algorithm: Byzantine Consensus in Blockchain

TL;DR

This work tackles consensus in blockchain under Byzantine faults without fixed fault bounds by introducing TWO-FOLD BFT (TDBA), a detection-based approach that uses per-node child processes to observe and compare messages against a parent process. The system builds a global blacklist of Byzantine nodes through cross-node sharing, allowing healthy nodes to ignore malicious traffic and continue consensus. Empirical simulations demonstrate detection rates exceeding $95\%$, with message complexity $O(n^2)$ and strong fault tolerance across large networks, outperforming several existing BFT and leaderless protocols. The method offers a practical path toward robust, scalable Byzantine consensus in asynchronous, permissionless blockchains by combining detection, blacklist propagation, and reliable broadcast without trusted authorities.

Abstract

Blockchain technology offers a decentralized and secure method for storing and authenticating data, rendering it well-suited for various applications such as digital currencies, supply chain management, and voting systems. However, the decentralized nature of blockchain also exposes it to vulnerabilities, particularly Byzantine faults, which arise when nodes in the network behave maliciously or encounter unexpected failures. Such incidents can result in inconsistencies within the blockchain and, in extreme scenarios, lead to a breakdown in consensus. Byzantine fault-tolerant consensus algorithms are crafted to tackle this challenge by ensuring that network nodes can agree on the blockchain's state even in the presence of faulty or malicious nodes. To bolster the system's resilience against these faults, it is imperative to detect them within the system. However, our examination of existing literature reveals a prevalent assumption: solutions typically operate under constraints regarding the number of faulty nodes. Such constraints confine the proposed solutions to ideal environments, limiting their practical applicability. In response, we propose a novel approach inspired by social paradigms, employing a trusted and fully monitored communication sub-process to detect Byzantine nodes. Upon detection, these nodes can be either disregarded in the consensus-building process, subjected to penalties, or undergo modifications as per the system's policy. Finally, we statistically demonstrate that our approach achieves a detection probability that exceeds 95\% for Byzantine nodes. In essence, our methodology ensures that if Byzantine nodes exhibit malicious behavior, healthy nodes can identify them with a confidence level of 95\%.

Figures (8)

• Figure 1: Message Structure
• Figure 2: Communication between Parent and Child
• Figure 3: Two-Fold Byzantine Flowchart
• Figure 4: Schema of proposed algorithm
• Figure 5: Comparison between Models (N = 100)