Blockchain-Aided Intrusion Detection in Marine Tactical Network Using Reinforcement Learning
Paper Title: Blockchain-Aided Intrusion Detection in Marine Tactical Network Using Reinforcement Learning
Journal Name: IEEE Transactions on Network and Service Management
Abstract: Marine tactical networks (MTNs) are essential for the safety and security of marine operations. Due to their sensitive nature, MTNs are susceptible to severe cyber risks, such as unauthorized access, data breaches, intrusion attacks. Traditional Intrusion Detection Systems (IDS) frequently have difficulties in effectively countering cyber threats in MTNs due to their static characteristics and restricted flexibility to changing threat environments. Thus, an IDS is vital for safeguarding MTNs from these vulnerabilities by proactively detecting and addressing threats. This paper proposes a novel blockchain-aided IDS for MTNs, leveraging reinforcement learning (RL) techniques. The proposed system aims to enhance the security and reliability of critical marine operations by effectively detecting and mitigating cyber threats. By integrating blockchain technology, the system ensures the integrity and immutability of intrusion detection data, preventing tampering and ensuring accountability. RL algorithms are employed to dynamically adapt the IDS's detection capabilities to evolving threat landscapes, improving its effectiveness and responsiveness. The key innovation of this proposed framework is integrating multi-agent adversarial RL, which enhances the model’s robustness and enables it to effectively identify and classify complex and dynamic threats within the MTNs. Experimental results demonstrate the superior performance of the proposed system in terms of detection accuracy, false positive rate, and computational efficiency compared to traditional IDS approaches. The proposed framework weighted accuracy of 80.16% and 95.9% for NSL-KDD and AWID datasets, respectively. This shows the robustness and applicability of the proposed model in complex MTN environments.