摘 要：随着物联网技术的快速发展，网络设备数量激增，传统入侵检测系统因计算资源消耗大而难以适应物联网环境的需求。为此，本研究提出一种面向物联网环境的轻量级网络入侵检测系统设计方法，旨在以较低的资源开销实现高效的安全防护。该系统基于深度学习模型优化与特征提取算法改进，采用轻量化神经网络结构以降低计算复杂度，并结合边缘计算技术将部分处理任务迁移至本地节点，从而减少数据传输延迟和带宽占用。实验结果表明，所提系统在保持较高检测准确率的同时显著降低了内存和处理器资源的使用量，相较于现有方案展现出更优的实时性和能效比。本研究的主要创新点在于通过模型压缩与边缘智能融合实现了资源受限场景下的高性能入侵检测，为物联网安全防护提供了可行的技术路径。
关键词：物联网安全；轻量级入侵检测；深度学习优化
Abstract：With the rapid development of Internet of Things (IoT) technology, the number of network devices has surged, making traditional intrusion detection systems (IDSs) less suitable for IoT environments due to their high consumption of computational resources. To address this issue, this study proposes a lightweight network intrusion detection system design method tailored for IoT environments, aiming to achieve efficient security protection with minimal resource overhead. The system integrates optimizations in deep learning models and improvements in feature extraction algorithms, employing a lightweight neural network architecture to reduce computational complexity. Furthermore, it leverages edge computing technology to migrate part of the processing tasks to local nodes, thereby minimizing data transmission latency and bandwidth usage. Experimental results demonstrate that the proposed system not only maintains a high detection accuracy but also significantly reduces memory and processor resource utilization, exhibiting superior real-time performance and energy efficiency compared to existing solutions. The primary innovation of this research lies in achieving high-performance intrusion detection in resource-constrained scenarios through model compression and the integration of edge intelligence, providing a feasible technical approach for IoT security protection.
引言 1
一、物联网环境下的安全需求分析 1
(一)物联网安全威胁概述 1
(二)轻量级检测系统必要性 2
(三)安全需求与技术挑战 2
二、轻量级入侵检测系统架构设计 3
(一)系统架构总体设计 3
(二)核心模块功能划分 3
(三)资源受限环境适配策略 4
三、数据采集与特征提取方法研究 4
(一)数据采集技术选择 4
(二)特征提取算法优化 5
(三)数据预处理与降维分析 5
四、入侵检测模型构建与性能评估 6
(一)检测模型算法设计 6
(二)实验环境与数据集构建 6
(三)性能评估与结果分析 7
结论 7
参考文献 9
致谢 9
关键词：物联网安全；轻量级入侵检测；深度学习优化
Abstract：With the rapid development of Internet of Things (IoT) technology, the number of network devices has surged, making traditional intrusion detection systems (IDSs) less suitable for IoT environments due to their high consumption of computational resources. To address this issue, this study proposes a lightweight network intrusion detection system design method tailored for IoT environments, aiming to achieve efficient security protection with minimal resource overhead. The system integrates optimizations in deep learning models and improvements in feature extraction algorithms, employing a lightweight neural network architecture to reduce computational complexity. Furthermore, it leverages edge computing technology to migrate part of the processing tasks to local nodes, thereby minimizing data transmission latency and bandwidth usage. Experimental results demonstrate that the proposed system not only maintains a high detection accuracy but also significantly reduces memory and processor resource utilization, exhibiting superior real-time performance and energy efficiency compared to existing solutions. The primary innovation of this research lies in achieving high-performance intrusion detection in resource-constrained scenarios through model compression and the integration of edge intelligence, providing a feasible technical approach for IoT security protection.
Keywords: Internet Of Things Security；Lightweight Intrusion Detection；Deep Learning Optimization
引言 1
一、物联网环境下的安全需求分析 1
(一)物联网安全威胁概述 1
(二)轻量级检测系统必要性 2
(三)安全需求与技术挑战 2
二、轻量级入侵检测系统架构设计 3
(一)系统架构总体设计 3
(二)核心模块功能划分 3
(三)资源受限环境适配策略 4
三、数据采集与特征提取方法研究 4
(一)数据采集技术选择 4
(二)特征提取算法优化 5
(三)数据预处理与降维分析 5
四、入侵检测模型构建与性能评估 6
(一)检测模型算法设计 6
(二)实验环境与数据集构建 6
(三)性能评估与结果分析 7
结论 7
参考文献 9
致谢 9
