Research on Intrusion Detection Method Based on Improved LSTNet
LI Baozhu
Abstract：With the proliferation of security vulnerabilities in Internet-of-Things(IoT)systems,conventional intrusion-de-tection approaches are increasingly inadequate for real-time and accurate detection of botnet attacks.Existing methods,when de-ployed on resource-constrained devices,struggle to capture the temporal context of periodic and bursty attacks,resulting in models that are both computationally intensive and prone to high false-positive rates.To address these limitations,this paper proposes a compact yet effective intrusion-detection framework specifically tailored for botnet attacks in IoT environments.First,this paper de-rives a concise representation of the N-BaIoT dataset through traffic-pattern analysis,time-series modeling,and centralizes feature extraction.Second,it designs a lightweight detection architecture that synergistically combines a pruned DenseNet with an en-hanced LSTNet,enabling rapid identification of botnet behaviors.Experimental results on the N-BaIoT dataset demonstrate a detec-tion accuracy of 95％,while maintaining high precision and recall across diverse operational scenarios.These findings substantiate the proposed method's efficacy in real-world deployments and its robustness to previously unseen attack patterns.
Keywords：intrusion detectionLSTNetDenseNet
Publication Date：2025-12-20
Online Publishing Date：2026-03-23(First online date of this platform, not the publication date of the document)
Pages：5( 124-128 )