Scholarlink

<p>The security of a data is the data is the prime challenge in the current scenario. To protect our data from unauthorized access we need a system that analyses traffic to identify network attacks, as well as alert the user by generating an alarm. In recent times numerous intrusion detection models have been built to detect intrusion activity. In this paper, we proposed the most effective hybrid Intrusion detection system (IDS) based on association rule mining (ARM) and ant colony optimization (ACO) for the identification of the attacks. The standard NSL-KDD dataset and CICIoT2023 dataset is considered for evaluating the performance of the proposed hybrid approach. The novel ARM-ACO hybrid model outperforms in detection accuracy as well as in the very low false alarm rate. The major advancement in the proposed model is that it also considered the categorical data which is not evaluated or considered in the previous methodology on unbiased selection of a dataset. As compared to earlier methodologies, our results were superior in terms of categorization as well as accuracy. In NSL-KDD dataset the accuracy of the proposed model is 99.80%, with a very low false alarm rate and in CICIoT2023 dataset the accuracy is 99.93% which is much better than the previous approaches. The proposed ARM-ACO model clearly shows its outperformance in terms of accuracy as well as low false alarm rate.</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 defense: Leveraging ant colony optimization for enhanced multi-optimization in network security

  • 期刊论文

  • 发表日期：

    2025年2月15日

查看全文
  • Chetan Gupta,
  • Amit Kumar,
  • Neelesh Kumar Jain

摘要

The security of a data is the data is the prime challenge in the current scenario. To protect our data from unauthorized access we need a system that analyses traffic to identify network attacks, as well as alert the user by generating an alarm. In recent times numerous intrusion detection models have been built to detect intrusion activity. In this paper, we proposed the most effective hybrid Intrusion detection system (IDS) based on association rule mining (ARM) and ant colony optimization (ACO) for the identification of the attacks. The standard NSL-KDD dataset and CICIoT2023 dataset is considered for evaluating the performance of the proposed hybrid approach. The novel ARM-ACO hybrid model outperforms in detection accuracy as well as in the very low false alarm rate. The major advancement in the proposed model is that it also considered the categorical data which is not evaluated or considered in the previous methodology on unbiased selection of a dataset. As compared to earlier methodologies, our results were superior in terms of categorization as well as accuracy. In NSL-KDD dataset the accuracy of the proposed model is 99.80%, with a very low false alarm rate and in CICIoT2023 dataset the accuracy is 99.93% which is much better than the previous approaches. The proposed ARM-ACO model clearly shows its outperformance in terms of accuracy as well as low false alarm rate.