Scholarlink

<p>Intrusion Detection Systems (IDS) are increasingly challenged by the complexity, dimensionality, and energy demands of modern network environments, where achieving high detection accuracy often comes at the cost of excessive computational overhead. Existing metaheuristic optimizers, though effective in feature selection and parameter tuning, frequently suffer from premature convergence and high evaluation costs, limiting their real-time applicability. To address these limitations, this paper proposes the Vampire Squid Optimization Algorithm (VSOA)—a novel bio-inspired metaheuristic that models the energy-efficient “drift and strike” foraging behavior of deep-sea vampire squids. VSOA introduces three innovations: (1) a dual-phase adaptive search mechanism that dynamically balances global exploration and local exploitation; (2) a bioluminescent threshold control for selective intensification based on contextual fitness; and (3) an energy-aware phase scheduler that minimizes redundant evaluations for computational efficiency. Applied to benchmark IDS datasets (NSL-KDD, UNSW-NB15, and CICIDS2017), VSOA demonstrates consistent superiority over state-of-the-art optimizers such as PSO, GA, Bat, and HFOA, achieving up to 3.8% higher detection accuracy and a 22% reduction in false-positive rate while using fewer fitness evaluations. These findings confirm that VSOA delivers a robust and resource-efficient optimization strategy for next-generation IDS, effectively bridging the gap between detection accuracy and energy-constrained deployment.</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

Vampire squid optimization algorithm for energy-efficient intrusion detection in cyber-physical networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月10日

查看全文
  • Ali Mohammed Alqaraghuli,
  • Abdullahi Ibrahim

摘要

Intrusion Detection Systems (IDS) are increasingly challenged by the complexity, dimensionality, and energy demands of modern network environments, where achieving high detection accuracy often comes at the cost of excessive computational overhead. Existing metaheuristic optimizers, though effective in feature selection and parameter tuning, frequently suffer from premature convergence and high evaluation costs, limiting their real-time applicability. To address these limitations, this paper proposes the Vampire Squid Optimization Algorithm (VSOA)—a novel bio-inspired metaheuristic that models the energy-efficient “drift and strike” foraging behavior of deep-sea vampire squids. VSOA introduces three innovations: (1) a dual-phase adaptive search mechanism that dynamically balances global exploration and local exploitation; (2) a bioluminescent threshold control for selective intensification based on contextual fitness; and (3) an energy-aware phase scheduler that minimizes redundant evaluations for computational efficiency. Applied to benchmark IDS datasets (NSL-KDD, UNSW-NB15, and CICIDS2017), VSOA demonstrates consistent superiority over state-of-the-art optimizers such as PSO, GA, Bat, and HFOA, achieving up to 3.8% higher detection accuracy and a 22% reduction in false-positive rate while using fewer fitness evaluations. These findings confirm that VSOA delivers a robust and resource-efficient optimization strategy for next-generation IDS, effectively bridging the gap between detection accuracy and energy-constrained deployment.