Stealthiness Analysis of the Identified Model-Based Covert Attacks with Unknown Model

Most  of  existing  covert  attacks  are  achieved  basedon known model, which is however impractical for the attacker,leading  to  the  failure  of  covert  attacks.  To  solve  this  problem,this paper presents a novel covert attacks method with unknownmodel.  Firstly,  the  negative  correlation  between  the  model  er-rors  and  the  stealthiness  of  traditional  covert  attacks  (TCAs)is  quantitatively  analyzed,  where  bigger  errors  brings  worsestealthiness. Secondly, the subspace identification method is usedto obtain the identified model of the plant. Furthermore, a novelidentification  and  adaptive  compensation  based  covert  attacks(IACCAs) method is proposed, which can adaptively compensatethe  attack  signals  by  adaptive  law.  Although  there  exist  theidentification errors, the stealthiness of IACCAs is proved to besimilar as TCAs with known model. Finally, experimental resultsdemonstrate  the  feasibility  and  effectiveness  of  the  proposedIACCAs method, which can achieve covert attacks with unknown model.