Reinforcement Learning Based Active Attack Detection and Blockchain Technique to Protect the Data from the Passive Attack in the Autonomous Mobile Network

【Author】 Sivasankar, C.; Kumanan, T.

【Source】WIRELESS PERSONAL COMMUNICATIONS

【影响因子】2.017

【Abstract】Nowadays, autonomous mobile network (AMN) nodes are utilized in widespread distribution; thus, security issues are a large apprehension. AMN is still necessary for further developments in terms of protection and privacy. In AMN, nodes are moved randomly, and the main concern is the absence of Infrastructure and dynamic topology. Active and passive attacker nodes are extremely damaging attacks alongside the AMN, and it builds the complete networks broken and disturbs the data transmission. To solve these issues, reinforcement learning based active attack detection and blockchain technique (RLBT) to protect the data from Passive attacks in the AMN. In this approach, we use the reinforcement learning (RL) method to detect the active attacks based on packet loss, packet deviation time, packet modification, and packet miss route. This blockchain method protects the node identity and other information. Thus, the passive attacker reads the data but can't identify the original data. As a result, the blockchain method provides secure communication. The RLBT approach main significant work in AMN is to isolate the active attacker nodes and protect the node information from the passive attacker nodes. Hence, this approach provides an important security service comprising authentication, privacy, and reliability. The experimental results demonstrate that the RLBT system increases 35% the detection ratio and 17.89% throughput in the AMN. Furthermore, it minimizes the possibility of miss-detection, and the false alarm rate is less than 0.1 in the AMN.

【Keywords】Active attack detection; Autonomous mobile network; Passive attack detection; Blockchain method; Reinforcement learning

【发表时间】2023 2023 JUL 19

【收录时间】2023-08-05

【文献类型】实验仿真

【主题类别】

区块链技术-协同技术-强化学习