ABSTRACT As a prevalent security threat to cyber‐physical systems, data tampering attacks deliberately alter transmitted measurements to degrade system identification performance. This paper studies the parameter identification problem of Finite Impulse Response (FIR) systems subjected to sequence data tampering attacks. First, sequence data tampering attacks are modeled based on binary observations, and the attack effects are systematically analyzed. Second, from the perspective of data defense, corresponding parameter identification algorithms are proposed for two typical attack scenarios (known and unknown attack frequency), including the definition of attack sequence identifiability, and the convergence and asymptotic normality of the algorithms are proved. Finally, the effectiveness of the proposed algorithms is validated through numerical simulations.
Li et al. (Tue,) studied this question.