网络安全与数据治理2023，Vol.42Issue(12)：48-53,6.DOI:10.19358/j.issn.2097-1788.2023.12.008
基于ADE-ABiGRU的物联网安全态势预测
Internet of Things security posture prediction based on ADE-ABiGRU
摘要
Abstract
Addressing the complexity and variability in IoT security situation prediction,this paper proposes an ADE-ABiGRU-based IoT security posture prediction model.The model merges bidirectional gated recurrent units,multi-head attention mecha-nisms,and residual structures,optimized through adaptive differential evolution to enhance its ability to capture complex temporal dependencies and analyze data across multiple dimensions.Refinement of the adaptive mechanism within the adaptive differential evolution algorithm ensures thorough consideration of temporal data characteristics,improving global search efficiency and local approximation accuracy.Experimental results on the ToN_IoT dataset show that the model outperforms traditional algorithms in terms of MAPE,R2,and MSE,demonstrating higher predictive accuracy and stability.关键词
网络安全/态势预测/双向门控循环单元/多头注意力机制/差分进化算法Key words
network security/posture prediction/bidirectional gated recurrent unit/multi-head attention mechanism/differenti-al evolution分类
信息技术与安全科学引用本文复制引用
彭兴维,袁凌云..基于ADE-ABiGRU的物联网安全态势预测[J].网络安全与数据治理,2023,42(12):48-53,6.基金项目
国家自然科学基金项目(62262073) （62262073）
云南省应用基础研究计划项目(202101AT070098) （202101AT070098）
云南省重大科技专项计划(202202AE090011) （202202AE090011）
