摘要
针对执行体资源受限场景下的系统威胁状态清除问题,提出一种异构资源受限的拟态裁决调度算法(LEAS)。该算法采用分阶段调度策略,结合随机调度、概率调度及推理调度3种方式,平衡调度开销与异常状态清除效率。实验结果表明,与同类应用场景下的两种常用调度算法相比,LEAS算法的平均异常调度周期在三者中最低,在清除系统威胁状态上具有更高的效率。
To address the system threat state removal under resource-constrained executors,a limited executer arbitration and scheduling algorithm(LEAS)is proposed.The phased scheduling strategy is developed by integrating random scheduling,probabilistic scheduling and inferential scheduling,achieving balance between scheduling overhead and abnormal state removal efficiency.Experimental results demonstrate that compared with other two commonly used scheduling algorithms in similar sce-narios,the proposed LEAS algorithm exhibits the shortest average anomaly scheduling cycle,suggest-ing superior efficiency in eliminating system threat states.
作者
于洪
葛晨洋
裴雪
兰巨龙
YU Hong;GE Chenyang;PEI Xue;LAN Julong(Information Engineering University,Zhengzhou 450001,China;Songshan Laboratory,Zhengzhou 450018,China)
出处
《信息工程大学学报》
2025年第3期290-296,共7页
Journal of Information Engineering University
基金
国家重点研发计划(2022YFB4401401)
河南省科技攻关计划(242102211066)。
关键词
动态异构冗余
拟态防御
内生安全
裁决
调度
dynamic heterogeneous redundancy
mimic defense
endogenous security
arbitration
scheduling
