Scholarlink

<p>In a data storage environment with high concurrency and multi-node collaboration, traditional intrusion detection methods still have limitations in parameter optimization, abnormal pattern recognition, and dynamic protection capabilities. To this end, this study proposes a network security data storage adaptive protection system that integrates the improved gray wolf optimization algorithm and support vector machine. First, it enhances the globality and stability of parameter search through Tent chaos initialization, nonlinear convergence factor and individual memory mechanism. Subsequently, the improved gray wolf optimization algorithm is used to optimize the key parameters of the support vector machine to improve the anomaly discrimination performance. Finally, combined with the situational awareness module, the risk intensity and threat trends are quantified, and threshold adaptive updates and real-time policy adjustments are realized. Experiments showed that on two types of public datasets, the recall rate of the proposed model increased to 98.2% and 98.7%, which was significantly better than the comparison models. In multi-scenario simulations, the system's security situation recognition accuracy remained above 93%, and the early warning time was increased by about 35–52% compared to the fixed strategy. In a multi-tenant scenario, the resource utilization balance reached a maximum of 0.97, showing good scalability and stability. Overall, the system has significant advantages in detection accuracy, policy adaptability, and robustness in complex environments, and provides a feasible solution for the security protection of large-scale network data storage environments.</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

Network security data storage adaptive protection system based on IGWO-SVM

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月18日

查看全文
  • Yixi Lei,
  • Hongkai Tian,
  • Yuan Du

摘要

In a data storage environment with high concurrency and multi-node collaboration, traditional intrusion detection methods still have limitations in parameter optimization, abnormal pattern recognition, and dynamic protection capabilities. To this end, this study proposes a network security data storage adaptive protection system that integrates the improved gray wolf optimization algorithm and support vector machine. First, it enhances the globality and stability of parameter search through Tent chaos initialization, nonlinear convergence factor and individual memory mechanism. Subsequently, the improved gray wolf optimization algorithm is used to optimize the key parameters of the support vector machine to improve the anomaly discrimination performance. Finally, combined with the situational awareness module, the risk intensity and threat trends are quantified, and threshold adaptive updates and real-time policy adjustments are realized. Experiments showed that on two types of public datasets, the recall rate of the proposed model increased to 98.2% and 98.7%, which was significantly better than the comparison models. In multi-scenario simulations, the system's security situation recognition accuracy remained above 93%, and the early warning time was increased by about 35–52% compared to the fixed strategy. In a multi-tenant scenario, the resource utilization balance reached a maximum of 0.97, showing good scalability and stability. Overall, the system has significant advantages in detection accuracy, policy adaptability, and robustness in complex environments, and provides a feasible solution for the security protection of large-scale network data storage environments.