Lateral movement detection

• US 9,825,978 B2
• Filed: 01/16/2017
• Issued: 11/21/2017
• Est. Priority Date: 09/18/2014
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Siva Kumar, Ram Shankar, Vu, Nguyen Song Khanh, DiPlacido, Marco, Nair, Vinod, Das, Aniruddha, Swann, Matt, Selvaraj, Keerthi, Sellamanickam, Sundararajan
• Primary Examiner(s): ZHAO, DON GORDON
• Application Number:
  US15/407,127
• Publication Number:
  US 20170126717A1
• Time in Patent Office: 309 Days
• CPC Class Codes: H04L 63/083 using passwords cryptographic mechanisms or cryptographic arrangements for entity authentication using a predetermined code H04L9/3226
• CPC Class Codes: H04L 63/10 for controlling access to devices or network resources
• CPC Class Codes: H04L 63/102 Entity profiles
• CPC Class Codes: H04L 63/1416 Event detection, e.g. attack signature detection
• CPC Class Codes: H04L 63/1425 Traffic logging, e.g. anomaly detection

Abstract

Lateral movement detection may be performed by employing different detection models to score logon sessions. The different detection models may be implemented by and/or utilize counts computed from historical security event data. The different detection models may include probabilistic intrusion detection models for detecting compromised behavior based on logon behavior, a sequence of security events observed during a logon session, inter-event time between security events observed during a logon session, and/or an attempt to logon using explicit credentials. Scores for each logon session that are output by the different detection models may be combined to generate a ranking score for each logon session. A list of ranked alerts may be generated based on the ranking score for each logon session to identify compromised authorized accounts and/or compromised machines. An attack graph may be automatically generated based on compromised account-machine pairs to visually display probable paths of an attacker.

Claims

1. A computer-implemented method for performing network intrusion detection in a computer network having multiple computing devices, the method comprising:

accessing historical logon session data related to activities performed in connection with multiple authorized accounts during logon sessions on corresponding computing devices in the computer network, the historical logon session data including data representing security events triggered during each of the logon sessions in response to the corresponding authorized account accessing one of the computing devices in the computer network; and

based on the received historical logon session data, generating models each configured to output a probability value indicating whether one or more security events related to a new logon session in connection with one of the multiple authorized account are indicative of a compromised behavior, the generated models individually including a historical occurrence value of a distinct combination of the security events triggered during the individual logon sessions.

2. The computer-implemented method of claim 1 wherein the security events can include at least some of events of account logon/logoff, authentication, account management, process creation, process termination, directory service, object access, application initiation, application termination, file sharing, policy change, privileged use, or a system event.

3. The computer-implemented method of claim 1 wherein one of the distinct combinations of the security event variables is configured to assess whether a logon behavior is indicative of compromised behavior, the distinct combination of the security event variables include at least some of an account, an account type, a machine role, a machine role type, a time of logon, or a logon type.

4. The computer-implemented method of claim 1 wherein one of the distinct combinations of the security event variables is configured to assess whether a sequence of security events observed during the logon session is indicative of compromised behavior based on a historical occurrence value of the same sequence of security events observed during previous logon sessions.

5. The computer-implemented method of claim 4 wherein the historical occurrence value of the same sequence of security events during previous logon sessions includes a count of the same sequence of security events observed during previous logon sessions.

6. The computer-implemented method of claim 1 wherein one of the distinct combinations of the security event variables is configured to assess whether an inter-event time between security events is indicative of compromised behavior, the distinct combination of the security event variables include an account or account type and an associated sequence of security events and corresponding inter-event times between successive security events in the sequence.

7. The computer-implemented method of claim 1 wherein one of the distinct combinations of the security event variables is configured to assess whether an attempt to logon using explicit credentials during the logon session is indicative of compromised behavior, the distinct combination of the security event variables include at least some of an account, an account type, a machine role, a machine role type, a time of attempted logon related to logon attempt using explicit credentials.

8. The computer-implemented method of claim 1 wherein one of the distinct combinations of security event variables include at least one of:

a combination of at least some of an account, an account type, a machine role, a machine role type, a time of logon, or a logon type;

a combination of at least some of an account, an account type, a machine role, or a machine role type, and a sequence of security events;

a combination of an account or account type and an associated sequence of security events and corresponding inter-event times between successive security events in the sequence;

or

a combination of at least some of an account, an account type, a machine role, a machine role type, a time of attempted logon related to logon attempt using explicit credentials.

9. A computing device in a computer network containing multiple other computing devices, the computing device comprising:

a processor for executing computer-executable instructions; and

memory storing computer-executable instructions executable by the processor to cause the processor to;

access historical logon session data related to activities performed in connection with multiple authorized accounts during logon sessions on corresponding computing devices in the computer network, the historical logon session data including data representing security events triggered during each of the logon sessions in response to the corresponding authorized account accessing one of the computing devices in the computer network; and

based on the received historical logon session data, generate one or more models individually configured to output a probability value indicating whether one or more security events related to a new logon session in connection with one of the multiple authorized account are indicative of a compromised behavior, the generated one or more models individually including a historical occurrence value of a distinct combination of the security events triggered during the individual logon sessions.

10. The computing device of claim 9 wherein the security events can include at least some of events of account logon/logoff, authentication, account management, process creation, process termination, directory service, object access, application initiation, application termination, file sharing, policy change, privileged use, or a system event.

11. The computing device of claim 9 wherein one of the distinct combinations of the security event variables is configured to assess whether a logon behavior is indicative of compromised behavior, the distinct combination of the security event variables include at least some of an account, an account type, a machine role, a machine role type, a time of logon, or a logon type.

12. The computing device of claim 9 wherein one of the distinct combinations of the security event variables is configured to assess whether a sequence of security events observed during the logon session is indicative of compromised behavior based on a historical occurrence value of the same sequence of security events observed during previous logon sessions.

13. The computing device of claim 12 wherein the historical occurrence value of the same sequence of security events during previous logon sessions includes a count of the same sequence of security events observed during previous logon sessions.

14. The computing device of claim 9 wherein one of the distinct combinations of the security event variables is configured to assess whether an inter-event time between security events is indicative of compromised behavior, the distinct combination of the security event variables include an account or account type and an associated sequence of security events and corresponding inter-event times between successive security events in the sequence.

15. The computing device of claim 9 wherein one of the distinct combinations of the security event variables is configured to assess whether an attempt to logon using explicit credentials during the logon session is indicative of compromised behavior, the distinct combination of the security event variables include at least some of an account, an account type, a machine role, a machine role type, a time of attempted logon related to logon attempt using explicit credentials.

16. The computing device of claim 9 wherein one of the distinct combinations of security event variables include at least one of:

a combination of at least some of an account, an account type, a machine role, a machine role type, a time of logon, or a logon type;

a combination of at least some of an account, an account type, a machine role, or a machine role type, and a sequence of security events;

a combination of an account or account type and an associated sequence of security events and corresponding inter-event times between successive security events in the sequence;

or

a combination of at least some of an account, an account type, a machine role, a machine role type, a time of attempted logon related to logon attempt using explicit credentials.

17. A computer-implemented method for performing network intrusion detection in a computer network having multiple computing devices, the method comprising:

receiving logon session data of activities performed related to authorized accounts during logon sessions on a corresponding computing device in the computer network, the logon session data including data representing security events triggered during the logon sessions in response to the authorized accounts accessing the corresponding computing device in the computer network;

for the logon sessions;

deriving one or more probabilities of intrusion related to the individual logon sessions based on a comparison of the logon session data with distinct combinations of security event variables and a historical occurrence value corresponding to the individual distinct combinations of the security event variables, the individual probabilities of intrusion indicating whether one or more security events related to the individual logon sessions are indicative of a compromised behavior; and

combining the derived one or more probabilities of instruction into an overall probability related to the individual logon sessions; and

identifying one or more of compromised authorized accounts or compromised computing devices in the computer network based on the overall probabilities of the logon sessions related to the logon sessions by the authorized accounts on a corresponding computing device in the computer network.

18. The computer-implemented method of claim 17 wherein:

each overall probability corresponding to one of the logon sessions is assigned a weighting factor; and

combining the derived multiple probabilities includes combining the derived multiple probabilities using the weighting factors.

19. The computer-implemented method of claim 17 wherein the distinct combinations of security event variables include at least some of:

a combination of at least some of an account, an account type, a machine role, a machine role type, a time of logon, or a logon type;

a combination of at least some of an account, an account type, a machine role, or a machine role type, and a sequence of security events;

a combination of an account or account type and an associated sequence of security events and corresponding inter-event times between successive security events in the sequence;

or

a combination of at least some of an account, an account type, a machine role, a machine role type, a time of attempted logon related to logon attempt using explicit credentials.

20. The computer-implemented method of claim 17 wherein one distinct combination of the security event variables is configured to assess whether a sequence of security events observed during the logon session is indicative of compromised behavior based on a historical occurrence value of the same sequence of security events observed during previous logon sessions.