计算机科学
正确性
地铁列车时刻表
数学优化
传输(电信)
最大化
协方差
无线传感器网络
集合(抽象数据类型)
实时计算
效用最大化问题
操作员(生物学)
解耦(概率)
随机过程
最优化问题
缩小
无线
调度(生产过程)
无线网络
稳健性(进化)
编码(社会科学)
分布式计算
协方差矩阵
作者
Yao Li,Jiawang Zhao,Zhichao Zhang,Xue Zhang,Xinping Guan
标识
DOI:10.1109/tsp.2025.3637549
摘要
In this paper, we investigate a power-constrained optimal denial-of-service (DoS) attack schedule against transmission over interactive wireless sensor networks. A novel scenario is considered in which an attacker aims to attack the transmission links of the interactive networks to maximize the global error covariance with constrained energy budget. The attacker decides whether to launch an attack or not at each time slot. Two different cost functions are considered, i.e., terminal-error case and average-error case. Through decoupling analysis, the global error covariance can be separated into two portions: one part generated by the initial error, and the other by system noise. Based on this separation, the analytical solution for the terminalerror case is clearly derived. For the average-error case, the indices “occurrence” and “concentration” are introduced. These two indices are used to quantitatively describe the number of times a certain operator appears in a set of error terms generated by iteration, and the accumulative effect of attacks over a time scale, respectively. Aided by these two proposed indices, the optimal attack schedule for the average-error case is analytically derived. Additionally, a problem with regard to the maximization ofoccurrenceis proposed, whose result can be proved to share the same form as the optimal attack schedule for average-error case with measurement noise. Moreover, we have extended our theoretical results to those of distributed scenarios, which follow the similar temporal distributions of global ones. Numerical simulations, including comparison with the learning-based state-of-the-art method, have demonstrated the correctness and effectiveness of our proposed scheduling strategies.
科研通智能强力驱动
Strongly Powered by AbleSci AI