Briefing
This dissertation addresses the critical absence of formal methodologies for rigorously evaluating the security of blockchain consensus algorithms, particularly their liveness in permissioned environments. It proposes a novel framework that employs state-change-based classification and a “security ingredients” taxonomy to systematically analyze consensus protocols. The foundational breakthrough lies in applying queueing theory and Markov chains to quantify the impact of malicious miners on system availability, revealing that while Lightweight Mining (LWM) demonstrates resilience, Byzantine Fault-Tolerant Raft (BFT-Raft) is critically vulnerable to even a single malicious actor. This new theoretical understanding provides a roadmap for designing more robust blockchain architectures, ensuring their operational continuity amidst adversarial conditions.
Context
Prior to this research, the development and customization of blockchain consensus algorithms often relied on developers’ experience and intuition, lacking a formal, mathematically rigorous approach to assess their security and performance. This prevailing theoretical limitation meant that the underlying reasons for a blockchain system’s operational correctness were not fully understood or formally proven. The absence of a systematic methodology to evaluate properties like liveness and safety against malicious behavior, such as denial-of-service attacks by miners, posed a significant challenge to establishing verifiable trust and reliability in distributed ledger technologies, particularly for permissioned blockchain systems.
Analysis
The paper’s core mechanism involves a multi-faceted methodology to evaluate consensus algorithm liveness. First, it introduces a state-change-based classification for Digital Ledger Technologies (DLTs) consensus algorithms, categorizing them into leader-based and voting-based types, with further sub-classifications. This framework facilitates a granular understanding of how miner-selection processes influence system state changes. Second, a “security ingredients” taxonomy is established, mapping prerequisites for achieving liveness, safety, and Byzantine Fault Tolerance (BFT).
The methodology then applies formal methods, specifically queueing theory and Markov chains, to model and simulate the impact of malicious miners on transaction waiting times and system availability. For example, in the Lightweight Mining (LWM) algorithm, the probability of good miners being selected and the average waiting time for transactions are calculated, demonstrating its resilience. Conversely, the analysis formally proves that BFT-Raft, without specific modifications, is highly susceptible to a single malicious miner, leading to system halts. This approach fundamentally differs from previous reliance on empirical observation by providing a theoretical and simulation-based quantification of resilience.
Parameters
- Core Concept ∞ Liveness Analysis Methodology
- New Classification ∞ State-Change-Based Consensus Algorithm Classification
- Security Framework ∞ Security Ingredients Taxonomy
- Formal Methods ∞ Queueing Theory, Markov Chains
- Evaluated Algorithms ∞ Lightweight Mining (LWM), BFT-Raft (Tangaroa)
- Threat Model ∞ Malicious Miner Denial-of-Service (DoS) Attacks
- System Focus ∞ Permissioned Blockchains
- Key Finding (BFT-Raft) ∞ Vulnerable to single malicious miner
- Key Finding (LWM) ∞ Resilient to high malicious miner count
- Authors ∞ Amani Altarawneh et al.