System and methods for automated computer security policy generation and anomaly detection
A method includes receiving first network traffic data having a plurality of data packets. A representation of each data packet from the plurality of data packets is stored in a first data store, to produce a set of training data. A plurality of communication protocols associated with the set of training data is detected. A set of rules based on the plurality of communication protocols is determined. The set of rules includes domain-specific information associated with each communication protocol detected. Training data from the set of training data is restructured according to the set of rules to produce processed data. The processed data is stored as non-malicious data in a second data store. A security policy is generated based on the processed data.
1 . A method, comprising:
receiving first network traffic data including a plurality of data packets;
storing a representation of each data packet from the plurality of data packets in a first data store, to produce a set of training data;
detecting a plurality of communication protocols associated with the set of training data;
determining a set of rules based on the plurality of communication protocols, wherein the set of rules includes domain-specific information associated with each communication protocol detected;
restructuring training data from the set of training data according to the set of rules, to produce processed data;
storing the processed data, as non-malicious data, in a second data store; and
generating a security policy based on the processed data.
2 . The method of claim 1 , further comprising:
receiving, via a user interface, a user input; and
generating the security policy in response to the user input.
3 . The method of claim 1 , further comprising:
storing the security policy in the second data store.
4 . The method of claim 1 , further comprising:
applying the security policy to second network traffic data.
5 . The method of claim 1 , wherein the plurality of communication protocols includes at least one of: Modbus, Financial Information exchange (FIX), Distributed Network Protocol (DNP3), Ethernet Global Data (EGD), Ethernet/IP, Message Queuing Telemetry Transport (MQTT), Constrained Application protocol (COAP), IEC104, OPC Unified Architecture (UA) controls, Niagara Fox, Factory Interface Network Service (FINS), or Seamless Message Protocol (SLMP).
6 . The method of claim 1 , wherein the first network traffic data includes known non-malicious data.
7 . The method of claim 1 , wherein the producing the set of training data is performed during a time period associated with known non-malicious data transmission.
8 . A method comprising:
receiving first network traffic data including a first plurality of data packets;
storing a representation of each data packet from the first plurality of data packets in a first data store, to produce a set of training data;
detecting a plurality of communication protocols associated with the set of training data;
determining a set of rules based on the plurality of communication protocols, wherein the set of rules includes domain-specific information associated with each communication protocol detected;
restructuring training data from the set of training data according to the set of rules, to produce processed data;
storing the processed data, as non-malicious data, in a second data store; and
performing anomaly detection based on the processed data.
9 . The method of claim 8 , wherein the performing of the anomaly detection comprises:
receiving second network traffic data including a second plurality of data packets;
comparing the second network traffic data to the processed data; and
determining a deviation in the second network traffic data from the processed data.
10 . The method of claim 9 , further comprising:
when the deviation is determined, blocking the second plurality of data packets of the second network traffic data.
11 . The method of claim 8 , wherein the plurality of communication protocols includes at least one of: Modbus, Financial Information exchange (FIX), Distributed Network Protocol (DNP3), Ethernet Global Data (EGD), Ethernet/IP, Message Queuing Telemetry Transport (MQTT), Constrained Application protocol (COAP), IEC104, OPC Unified Architecture (UA) controls, Niagara Fox, Factory Interface Network Service (FINS), or Seamless Message Protocol (SLMP).
12 . The method of claim 8 , wherein the restructuring the training data from the set of training data according to the set of rules includes compressing the training data from the set of training data according to the set of rules.
13 . The method of claim 8 , wherein the restructuring of the training data includes normalizing the training data according to the set of rules.
14 . A method, comprising:
receiving first network traffic data including a plurality of data packets;
storing a representation of each data packet from the plurality of data packets in a first data store, to produce a set of training data;
detecting a plurality of communication protocols associated with the set of training data;
determining a set of rules based on the plurality of communication protocols, wherein the set of rules includes domain-specific information associated with each communication protocol detected;
restructuring training data from the set of training data according to the set of rules based on the detected plurality of communication protocols, to produce processed data;
storing the processed data, as non-malicious data, in a second data store; and
modifying the restructured data.
15 . The method of claim 14 , wherein the modifying of the restructured data comprises:
removing unwanted data from the processed data.
16 . The method of claim 14 , wherein the modifying of the restructured data comprises:
adding known good data to the processed data.
17 . The method of claim 14 , wherein the modifying of the restructured data comprises:
receiving the first network traffic data;
detecting known good data in the first network traffic data; and
adding the known good data to the processed data.
18 . The method of claim 14 , wherein the plurality of communication protocols includes at least one of: Modbus, Financial Information exchange (FIX), Distributed Network Protocol (DNP3), Ethernet Global Data (EGD), Ethernet/IP, Message Queuing Telemetry Transport (MQTT), Constrained Application protocol (COAP), IEC104, OPC Unified Architecture (UA) controls, Niagara Fox, Factory Interface Network Service (FINS), or Seamless Message Protocol (SLMP).
19 . The method of claim 14 , wherein the first network traffic data includes known non-malicious data.
20 . The method of claim 14 , wherein the restructuring of the training data includes compressing the training data.