异步拒绝服务攻击下神经网络的安全同步控制OA
Secure Synchronization Control of Neural Networks Under Asynchronous DoS Attacks
针对一类神经网络同步系统,当考虑神经网络不同通道受到拒绝服务(denial of service,DoS)攻击时,提出一种双通道动态事件触发机制策略.首先,考虑传感器到控制器和控制器到执行器受到不同DoS攻击的情况,并将这些攻击建模为2个独立的马尔可夫(Markov)跳变过程;其次,为了应对数据传输拥塞问题,设计了双通道动态事件触发机制来提高传输效率;然后,为了满足同步要求,提出了相应的控制器设计;最后,利用线性矩阵不等式的解析方法,验证了所提理论的可行性,并通过实验结果证明了该方法的优越性.
This paper considers a class of synchronization systems about neural networks and proposes a dual dynamic event-triggered mechanism strategy when different channels are subjected to denial of serv-ice(DoS)attacks.First,the article considers the case where the sensor to controller and controller to ac-tuator are subjected to different DoS attacks and model these attacks as two independent Markov jump processes.Second,in order to cope with data transmission congestion,the article introduces two event-trig-gered mechanisms to improve the transmission efficiency.Then,to meet the synchronization requirements,the article proposes a corresponding controller design.Finally,by using the analytical method of linear ma-trix inequalities,the feasibility of the proposed theory is verified,and the superiority of the method is veri-fied through experimental results.
杨邹;施开波;王俊;朱文卿
西南民族大学 电气工程学院,四川成都 610041成都大学电子信息与电气工程学院,四川成都 610106西南民族大学 电气工程学院,四川成都 610041西南民族大学 电气工程学院,四川成都 610041
信息技术与安全科学
神经网络Markov跳变过程异步DoS攻击动态事件触发机制同步控制
neural networksMarkov jump processesasynchronous DoS attacksdynamic event-triggered mechanismsynchronization control
《成都大学学报（自然科学版）》 2026 (1)
39-44,6
国家自然科学基金项目(61703060、12061088)四川省科技厅重点研发计划项目(2023YFG0287)四川省自然科学青年基金项目(24NSFSC7038)
评论