DDoS Attack Detection in Business Blockchain Networks: A Review, Framework, and Challenges

With the rapid growth of the Internet of Things and the increasing deployment of blockchain technology in critical applications, distributed denial of service (DDoS) attacks pose a significant threat to the security and availability of blockchain networks. However, due to the lack of data sharing among different business chains, existing attack detection systems fail to identify global DDoS attack patterns. This leads to reduced detection accuracy and timeliness, as well as increased risks of false positives and false negatives. Moreover, it prevents the implementation of distributed collaborative defense mechanisms, allowing attackers to bypass local defenses through cross-chain attacks.This paper reviews the current state of DDoS attack detection in blockchain networks and provides a preliminary look at the challenges facing the field.To address these challenges, this paper proposes a new intrusion detection system, BAFL, which utilizes supervision chain and federated learning to detect DDoS attacks in blockchain network layer. We design a distributed DDoS architecture that utilizes a supervised chain of directed acyclic graph (DAG) structures to achieve trusted sharing of model parameters among client nodes and record detected DDoS attacks. The system uses asynchronous federated learning algorithm, weights the client model based on its DDoS attack detection accuracy, and introduces a time attenuation factor to reduce the influence of old model parameters on aggregation and accelerate model convergence. Through comprehensive experiments on data sets such as CIC-DDoS2019, we demonstrated that the accuracy of BAFL in detecting DDoS attacks is more than 99%, which is better than traditional methods, the accuracy is increased by 2%, and the real-time detection delay is reduced by half. The results highlight the robustness and scalability of BAFL, making it a promising solution for protecting blockchain networks from DDoS threats.