Assessing the Resilience of BFT-Raft Consensus Against Insider DoS Attacks in Blockchain
Amani Altarawneh, Jemima Owusu-Tweneboah · 2025
A significant amount of research has put forward various consensus mechanisms for blockchain technology, but many have not been implemented due to security and performance concerns. Consensus mechanisms are essential for ensuring the continued operation (liveness) and security of blockchain systems. The Byzantine Fault Tolerant-Raft (BFT-Raft) consensus algorithm combines the simplicity of Raft, developed by the Stanford group, with the resilience of Byzantine Fault Tolerance (BFT) mechanisms. However, the liveness and security of BFT-Raft as a consensus mechanism have not been thoroughly validated. This study addresses this gap by analyzing the availability of BFT-Raft, particularly focusing on the behavior of malicious miners who manipulate their timeout to influence the leader election process within the consensus mechanism. Using both simulation and theoretical models, the research assesses the vulnerability of BFT-Raft to insider denial of service (DoS) attacks. Additionally, the study evaluates BFT-Rafts' ability to maintain availability and effectively terminate the consensus and miner selection processes in the presence of malicious miners and clients. Using binomial distribution, queuing theory, and Markov chains, the resilience of BFT-Raft is systematically examined. The findings indicate that BFT-Raft has specific security vulnerabilities making it susceptible to insider DoS attacks. The paper also proposes potential approaches to mitigate these vulnerabilities, such as a game theory approach with incentives and penalties, a hardware-based approach, and an attack detection approach using AI and explainable AI to help make decisions for these complex attacks.