Attack Detection and Identification in Cyber-Physical Systems
read more
Citations
Input-to-State Stabilizing Control Under Denial-of-Service
Survey of intrusion detection systems: techniques, datasets and challenges
A survey on security control and attack detection for industrial cyber-physical systems
Networked control systems: a survey of trends and techniques
Physical Authentication of Control Systems: Designing Watermarked Control Inputs to Detect Counterfeit Sensor Outputs
References
Decoding by linear programming
MATPOWER: Steady-State Operations, Planning, and Analysis Tools for Power Systems Research and Education
The IEEE Reliability Test System-1996. A report prepared by the Reliability Test System Task Force of the Application of Probability Methods Subcommittee
Power System State Estimation : Theory and Implementation
False data injection attacks against state estimation in electric power grids
Related Papers (5)
Secure Estimation and Control for Cyber-Physical Systems Under Adversarial Attacks
A secure control framework for resource-limited adversaries
Frequently Asked Questions (15)
Q2. What are the future works in "Attack detection and identification in cyber-physical systems" ?
Future and ongoing work includes ( i ) a detailed analysis of the convergence of their distributed monitor, ( ii ) the design of distributed identification monitors, and ( iii ) the design of monitors robust to system noise and unmodeled dynamics.
Q3. What are the types of networks that are cyber-physical?
Mass transport networks are cyber-physical systems, such as gas transmission and distributionnetworks [32], large-scale process engineering plants [33], and water networks.
Q4. What is the sparsity pattern of the graph?
3. Let Gs = (V , E) be the directed sparsity graph associated with the pair (E,A), where the vertex set V = X corresponds to the system state, and the set of directed edges E = {(xj, xi) : eij 6= 0 or aij 6= 0} is induced by the sparsity pattern of E and A.
Q5. What is the proof of Lemma 3.2?
For a linear descriptor system with smooth input and consistent initialcondition, the existence of zero dynamics is equivalent to the existence of invariant zeros asin (ii), see [28, Theorem 3.2 and Proposition 3.4].
Q6. What is the simplest way to show the error dynamics in the absence of attacks?
To show stability of the error dynamics in the absence of attacks, the authors employ the small-gain approach to large-scale systems and rewrite the error dynamics (12) as the closed-loopinterconnection of the two subsystems Γ1 : Eė(t) = (AD + GC)e(t) + v(t) and Γ2 : v(t) = ACe(t).
Q7. What is the proof of the structural left-invertibility?
Theorem 3.5 extends the structural left-invertibility results known for nonsingular systems toregular descriptor systems, and its proof relies on classical concepts from structural analysis,algebraic geometry, and graph theory.
Q8. What is the second equation of (7)?
Since the initial condition x(0) and the input uK are assumed to be consistent (A2) and non-impulsive (A3), the error system (6) has no invariant zeros if and only if [28, Proposition 3.4] there exists no triple (s, w̄, gK) ∈ C× Rn × Rp satisfying sE − (A+GC) BK +GDKC −DK w̄ gK = 0 0 . (7)The second equation of (7) yields Cw̄ = DKgK .
Q9. What is the i-th subgraph of Gs with vertices Vi and?
Let V be partitioned into N disjoint subsets as V = V1 ∪ · · · ∪ VN , with |Vi| = ni, and let Gis = (Vi, Ei) be the i-th subgraph of Gs with vertices Vi and edges Ei = E ∩ (Vi × Vi).
Q10. What is the structure of the matrix s[E]A?
the matrix s[E]−[A] is structurally non-degenerate if the determinant |sE−A| 6= 0 for a generic realization of E and A, that is, |sE −A| 6= 0 holds in the whole parameter space of elements of E and A with exception of a low dimensional variety [24], [38].
Q11. What is the third and final design step of the attack identification filter?
For the ease of notation and without affecting generality, the third and final design step ofour attack identification filter is presented for the pre-conditioned system (25).
Q12. What is the centralized attack identification procedure?
The following centralized attack identification procedure consists of designing a residual filter to determine whether a predefined set coincides with theattack set.
Q13. What is the vulnerability of open channel networks to cyber-physical attacks?
The vulnerability of open channel networks to cyber-physical attacks hasbeen studied in [12], [22], and municipal water networks are also known to be susceptible toattacks on the hydraulics [1] and biochemical contamination threats [23].
Q14. What is the simplest example of a cyber-physical system?
Remark 1: (Examples of cyber-physical systems requiring advanced security mechanisms)Future power grids will combine physical dynamics with a sophisticated coordination infrastruc-ture.
Q15. What is the simplest way to determine the asymmetrical detection of attacks?
Lemma 4.2: (Decentralized stabilization of the attack detection filter) Consider the descriptorsystem (1), and assume that the attack set K is detectable and that the network initial state x(0)is known.