| 伏金娣*,刘小杰**,宋长新**.基于并行优化与模糊聚类的入侵检测[J].高技术通讯(中文),2026,36(4):423~429 |
| 基于并行优化与模糊聚类的入侵检测 |
| Intrusion detection based on parallel optimization and fuzzy clustering |
| |
| DOI：10. 3772/ j. issn. 1002 -0470. 2026. 04. 009 |
| 中文关键词: 传感器云安全； 智能入侵检测； 并行特征选择； 离散优化； 自适应模糊聚类 |
| 英文关键词: sensor cloud security, intelligent intrusion detection, parallel feature selection, discrete optimization, adaptive fuzzy clustering |
| 基金项目: |
| 作者 | 单位 | | 伏金娣* | (*浙江理工大学科技与艺术学院绍兴 312369)
(**上海城建职业学院人工智能应用学院上海 201415) | | 刘小杰** | | | 宋长新** | |
|
| 摘要点击次数: 32 |
| 全文下载次数: 25 |
| 中文摘要: |
| 本文提出一种面向传感器云环境的入侵检测框架，融合了并行离散优化与机器学习技术，以提升系统安全性。首先，构建了最优特征评价准则，并设计并行离散优化特征提取系统，通过并行筛选机制有效降低数据维度并增强特征稳定性。其次，在离散优化过程中引入智能迭代进化策略，所开发的算法具有全局收敛性，能够高效获取最优特征子集。最后，结合自调节聚类方法对提取的特征进行分布式模糊聚类分析，该方法能自动确定最优聚类数目，并克服传统模糊聚类易陷入局部最优的问题，从而实现对入侵行为的精准识别。实验结果表明，所提算法提升了入侵判定的准确性，大幅降低了漏检率，且在含噪环境中仍保持稳定可靠的检测性能，展现出良好鲁棒性。 |
| 英文摘要: |
| This paper proposes an intrusion detection framework for sensor-cloud environments that integrates parallel discrete optimization with machine learning to enhance system security. Firstly, an optimal feature evaluation criterion is established, and a parallel discrete optimization-based feature extraction system is developed to reduce dimensionality and improve feature stability. Secondly, an intelligent iterative evolutionary strategy with global convergence is incorporated to efficiently obtain the optimal feature subset. Finally, a self-adaptive distributed fuzzy clustering method is employed to analyze the extracted features, automatically determining the number of clusters and alleviating local optima, thereby enabling accurate intrusion detection. Experimental results show that the proposed method achieves higher detection accuracy, a lower missed detection rate, and stable performance in noisy environments, demonstrating its strong robustness. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
