Blockchain Validator Node Intrusion Detection System Based On IRLG-S Model

The introduction of Web 3.0 concept has led to the wider application of blockchain technology.With the increasing number of blockchain users,the risk of attacks on blockchain nodes is gradually increasing,so network intrusion detection for blockchain verifier nodes is important,and many scholars have proposed various methods to deal with network intrusion.However,all of these methods have problems such as low model accuracy and no restriction on the management authority of blacklisted contracts.In this paper,we propose a network intrusion detection model based on IRLG-S.The model is improved on the basis of Inception-Res Net,and an LSTM-GRU model using residual structure connection is added to extract the temporal features of the data.For the CICIDS2017 dataset,three data balancing methods are used for comparison experiments;a deep learning and smart contract-based network intrusion detection system for blockchain verifier nodes is designed to store attacker IPs on the chain by using smart contracts to achieve that the entire blockchain network verifier nodes can obtain the attacker IPs and prevent the attackers from continue the attack.At the same time,the system will also capture the network traffic data of the visiting nodes and send them to the IRLG-S model to detect the presence of attacks,and upload the detected attacker IPs to the chain to achieve the continuous updating of the IP blacklist list.The IRLG-S model proposed in this paper achieves a high accuracy rate on the CICIDS2017 dataset,while the accuracy rate,recall rate,and F1 metrics are improved compared to recent studies.The system designed in this paper,with all functional modules meeting the expected functions,has good protection against blockchain verifier node network intrusion attacks.