We study the problem of identifying Finite Impulse Response (FIR) systems against random replay attacks with binary-valued observations in this paper. Replay attacks are modeled and the impact of attack strategies on the performance of parameter estimation algorithms is investigated. A defense algorithm that is consistent even during attacks is designed, and the problem of identifiability for the unknown parameters is discussed. Then, the asymptotic normality of this algorithm is given, and we derive the optimal defense strategy based on this. The proposed method’s rationality is verified by simulation example.

中文翻译：

---

在重放攻击下识别具有二进制值观测值的 FIR 系统


在本文中，我们研究了用二进制值观察来识别有限脉冲响应 （FIR） 系统免受随机重放攻击的问题。对重放攻击进行建模，并研究了攻击策略对参数估计算法性能的影响。设计了一种即使在攻击中也能保持一致的防御算法，并讨论了未知参数的可识别性问题。然后，给出该算法的渐近正态性，并以此推导最优防御策略。通过仿真算例验证了所提方法的合理性。