Industrial process system threat detection
Examples of techniques for threat detection in an industrial process system are described herein. An aspect includes determining a plurality of subsystems of an industrial process system. Another aspect includes, for each of the plurality of subsystems, constructing and training a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system. Another aspect includes monitoring the industrial process system using the plurality of DAE models corresponding to the plurality of subsystems. Another aspect includes, based on the plurality of DAE models, determining a cyberattack in a subsystem of the plurality of subsystems.
1. A system, comprising a processor configured to:
determine a plurality of subsystems of an industrial process system;
for each of the plurality of subsystems, construct and train a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system;
monitor the industrial process system uning the plurality of DAE models corresponding to the plurality of subsystems; and
based on the plurality of DAE models, determine a cyberattack in a subsystem of the plurality of subsystems,
wherein determining the plurality of subsystems of the industrial process system comprises:
determining a node graph of the industrial process system, the node graph comprising a plurality of nodes, each of the plurality of nodes corresponding to a respective element of the industrial process system;
determining a plurality of node clusters in the node graph, wherein each node cluster of the plurality of node clusters comprises a respective subset of the plurality of nodes, and wherein each node cluster of the plurality of node clusters corresponds to a respective subsystem of the industrial process system.
2. The system of claim 1 , wherein the plurality of nodes corresponds to information technology (IT) elements and operational technology (OT) elements of the industrial process system.
3. The system of claim 1 , wherein a DAE model of the plurality of DAE models is constructed based on elements of the industrial process system that correspond to the subset of the plurality of nodes in the node cluster that corresponds to the DAE model.
4. The system of claim 3 , wherein the DAE model receives input data corresponding to each of the elements that correspond to the subset of the plurality of nodes during the monitoring of the industrial process system.
5. The system of claim 1 , wherein determining the cyberattack in the subsystem comprises:
determining an error between input data of the DAE model corresponding to the subsystem and output data of the DAE model corresponding to the subsystem;
determining whether the error is above a threshold; and
based on the error being above the threshold, determining the cyberattack in the subsystem.
6. The system of claim 1 , further comprising identifying a type of the determined cyberattack based on the monitoring of the industrial process system.
7. A computer-implemented method, comprising:
determining, by a processor, a plurality of subsystems of an industrial process system;
for each of the plurality of subsystems, constructing and training, by the processor, a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system;
monitoring the industrial process system using the plurality of DAE models corresponding to the plurality of subsystems; and
based on the plurality of DAE models, determine a cyberattack in a subsystem of the plurality of subsystems,
wherein determining the plurality of subsystems of the industrial process system comprises:
determining a node graph of the industrial process system, the node graph comprising a plurality of nodes, each of the plurality of nodes corresponding to a respective element of the industrial process system;
determining a plurality of node clusters in node graph, wherein each node cluster of the plurality of node clusters comprises a respective subset of the plurality of nodes, and wherein each node cluster of the plurality of node clusters corresponds to a respective subsystem of the industrial process system.
8. The computer-implemented method of claim 7 , wherein the plurality of nodes corresponds to information technology (IT) elements and operational technology (OT) elements of the industrial process system.
9. The computer-implemented method of claim 7 , wherein a DAE model of the plurality of DAE models is constructed based on elements of the industrial process system that correspond to the subset of the plurality of nodes in the node cluster that corresponds to the DAE model.
10. The computer-implemented method of claim 9 , wherein the DAE model receives input data corresponding to each of the elements that correspond to the subset of the plurality of nodes during the monitoring of the industrial process system.
11. The computer-implemented method of claim 7 , wherein determining the cyberattack in the subsystem comprises:
determining an error between input data of the DAE model corresponding to the subsystem and output data of the DAE model corresponding to the subsystem;
determining whether the error is above a threshold; and
based on the error being above the threshold, determining the cyberattack in the subsystem.
12. The computer-implemented method of claim 7 , further comprising identifying a type of the determined cyberattack based on the monitoring of the industrial process system.
13. A computer program product comprising:
a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processinh device to cause the processing device to perform a method comprising:
determining a plurality of subsystems of an industrial process system;
for each of the plurality of subsystems, constructing and training a respective deep autoencoder (DAE) model of the subsystem based on data corresponding to the industrial process system;
monitoring the industrial process system using the plurality of DAE models corresponding to the plurality of subsystems; and
based on the plurality of DAE models, determine a cyberattack in a subsystem of the plurality of subsystems,
wherein determining the plurality of subsystems of the industrial process system comprises:
determining a node graph of the industrial process system, the node graph comprising a plurality of nodes, each of the plurality of nodes corresponding to a respective element of the industrial process system;
determining a plurality of node clusters in the node graph, wherein each node cluster of the plurality of node clusters comprises a respective subset of the plurality of nodes, and wherein each node cluster of the plurality of node clusters corresponds to a respective subsystem of the industrial process system.
14. The computer program product of claim 13 , wherein the plurality of nodes corresponds to information technology (IT) elements and operational technology (OT) elements of the industrial process system.
15. The computer program product of claim 13 , wherein a DAE model of the plurality of DAE models is constructed based on elements of the industrial process system that correspond to the subset of the plurality of nodes in the node cluster that corresponds to the DAE model.
16. The computer program product of claim 15 , wherein the DAE model receives input data corresponding to each of the elements that correspond to the subset of the plurality of nodes during the monitoring of the industrial process system.
17. The computer program product of claim 13 , wherein determining the cyberattack in the subsystem comprises:
determining an error between input data of the DAE model corresponding to the subsystem and output data of the DAE model corresponding to the subsystem;
determining whether the error is above a threshold; and
based on the error being above the threshold, determining the cyberattack in the subsystem.