Scholarlink

<p>The rapid growth of the internet and the increase of computer attacks have created challenges for today's networks. For this reason, intrusion detection systems (IDS) provide effective accuracy and adaptability for ever-changing, varied traffic patterns. Existing models struggle with generalization and adapting to dynamic, high-dimensional network traffic. Therefore, the research aims to develop an adaptable and accurate IDS using a hybrid method. A Monarch Butterfly Optimized Modified Gaussian Process Mixture (MBO-MGPM) method is suggested for implementing IDS. The MBO-MGPM method combines the MGPM model with MBO to produce a superior method for selecting model hyperparameters, stabilizing convergence in the MGPM model, and increasing model generalization ability. The NSL-KDD dataset with 125,973 records is used to create a model of normal traffic data. The collected data are pre-processed using steps including cleaning data, encoding categorical features into numerical format using one-hot encoding, and using min–max normalization. The MBO-MGPM technique, implemented in Python 3.8, outperforms conventional machine learning (ML) methodologies with better accuracy of 98.8%, coupled with a lower error rate and execution time. Moreover, by optimizing MBO parameters for the MGPM results in an increased level of classification accuracy over existing methods. Overall, this research supports the idea that MBO-MGPM could evolve into a workable methodology for addressing both intrusions and their corresponding limitations as experienced by current-day technologies. The MBO-MPGM approach would provide the logical step in supporting the defense of complex networks against cybercrime.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intrusion Detection in Wireless Sensor Networks Using a Machine Learning-Driven Cyber Security Framework

  • 期刊论文

  • 发表日期：

    2026年4月4日

查看全文
  • M. Meenakshi,
  • M. Mageshwari,
  • S. P. M. Kannan,
  • P. Veeramanikandan,
  • K. Manivannan,
  • Elangovan Muniyandy,
  • R. Mohan Kumar

摘要

The rapid growth of the internet and the increase of computer attacks have created challenges for today's networks. For this reason, intrusion detection systems (IDS) provide effective accuracy and adaptability for ever-changing, varied traffic patterns. Existing models struggle with generalization and adapting to dynamic, high-dimensional network traffic. Therefore, the research aims to develop an adaptable and accurate IDS using a hybrid method. A Monarch Butterfly Optimized Modified Gaussian Process Mixture (MBO-MGPM) method is suggested for implementing IDS. The MBO-MGPM method combines the MGPM model with MBO to produce a superior method for selecting model hyperparameters, stabilizing convergence in the MGPM model, and increasing model generalization ability. The NSL-KDD dataset with 125,973 records is used to create a model of normal traffic data. The collected data are pre-processed using steps including cleaning data, encoding categorical features into numerical format using one-hot encoding, and using min–max normalization. The MBO-MGPM technique, implemented in Python 3.8, outperforms conventional machine learning (ML) methodologies with better accuracy of 98.8%, coupled with a lower error rate and execution time. Moreover, by optimizing MBO parameters for the MGPM results in an increased level of classification accuracy over existing methods. Overall, this research supports the idea that MBO-MGPM could evolve into a workable methodology for addressing both intrusions and their corresponding limitations as experienced by current-day technologies. The MBO-MPGM approach would provide the logical step in supporting the defense of complex networks against cybercrime.