Malicious lateral movement detection using remote system protocols
A system and method for malicious lateral movement detection. A method includes identifying atomic tunnels in packets sent between devices; identifying tunnel constructs; determining a potentially malicious atomic tunnel among the atomic tunnels by comparing edges of each of the atomic tunnels to edges of previously observed tunnel constructs; determining a potentially malicious tunnel including the potentially malicious atomic tunnel; and mitigating the potentially malicious tunnel. Each atomic tunnel is a structure representing communications among the devices defined with respect to at least three nodes and at least two edges. Each node represents a respective device, and each edge represents a connection between two of the devices. Each atomic tunnel has two hops, where each hop is a level of communication in which a packet is sent from one device to another device. Each tunnel construct is a structure including at least one of the atomic tunnels.
1 . A method for malicious lateral movement detection, comprising:
identifying a plurality of atomic tunnels based on data including packets sent between devices among a plurality of devices, wherein each atomic tunnel is a structure representing communications among the plurality of devices defined with respect to at least three nodes and at least two edges, each node representing a respective device of the plurality of devices, each edge representing a connection between two of the plurality of devices, each atomic tunnel having two hops, wherein each hop is a level of communication in which a packet is sent from one device to another device, and wherein at least one device of the plurality of devices comprises a remotely accessed device accessed by one or more other devices of the plurality of devices using one or more remote system protocols;
identifying a plurality of tunnel constructs, wherein each tunnel construct is a structure including at least one of the plurality of atomic tunnels;
determining a potentially malicious atomic tunnel among the plurality of atomic tunnels by comparing the edges of each of the plurality of atomic tunnels to a plurality of edges of a plurality of previously observed tunnel constructs;
determining a potentially malicious tunnel, wherein the potentially malicious tunnel includes the potentially malicious atomic tunnel, and wherein determining the potentially malicious tunnel comprises:
performing a classification of the plurality of previously observed tunnel constructs to classify each previously observed tunnel construct into a class of a plurality of classes, wherein each class comprises an associated structure;
isomorphically comparing a tunnel structure of the potentially malicious atomic tunnel to the associated structure of each class of the plurality of classes; and
matching the tunnel structure with one of the associated structures to determine the potentially malicious tunnel; and
performing at least one mitigation action based on the determined potentially malicious tunnel.
2 . The method of claim 1 , wherein the potentially malicious atomic tunnel is one of the plurality of atomic tunnels for which at least one of the at least two edges of the potentially malicious atomic tunnel is an outlier edge as compared to the edges of the plurality of previously observed tunnel constructs.
3 . The method of claim 2 , further comprising:
analyzing session data for the edges of at least one tunnel construct of the plurality of tunnel constructs in order to determine whether each edge of the analyzed at least one tunnel construct is an outlier edge as compared to the edges of the plurality of previously observed tunnel constructs.
4 . The method of claim 3 , wherein the analyzed session data includes at least one of: time of beginning of session, total session duration, amount of incoming data, and amount of outgoing data.
5 . The method of claim 2 , wherein each edge of the potentially malicious atomic tunnel is either a new edge or an outlier edge as compared to the edges of the plurality of previously observed tunnel constructs, wherein a new edge is an edge that was not included among the plurality of previously observed tunnel constructs.
6 . The method of claim 1 , further comprising:
comparing a structure of the potentially malicious tunnel construct to a structure associated with each of a plurality of tunnel classes in order to determine a classification for the potentially malicious tunnel, wherein each of the plurality of tunnel classes is defined based on a respective portion of the plurality of previously observed tunnel constructs;
testing a first edge of the potentially malicious tunnel construct in order to determine whether the tested first edge is an outlier edge as compared to a first edge of each of at least one matching class tunnel construct of the plurality of previously observed tunnel constructs, wherein each of the at least one matching class tunnel construct has a same classification as the potentially malicious tunnel construct; and
detecting the potentially malicious tunnel as a malicious tunnel when the tested first edge is an outlier edge as compared to the first edge of each of the at least one matching class tunnel construct.
7 . The method of claim 1 , wherein, for each atomic tunnel, a connection time of a first hop of the atomic tunnel includes a connection time of a second hop of the atomic tunnel.
8 . The method of claim 1 , wherein, for each atomic tunnel, an amount of data sent in a first hop of the atomic tunnel includes an amount of data sent in a second hop of the atomic tunnel.
9 . The method of claim 1 , wherein performing the at least one mitigation action includes generating an alert including a risk level, wherein the risk level is determined based on at least one of: a number of devices represented by the nodes of the potentially malicious tunnel, and a number of potentially malicious atomic tunnels included in the potentially malicious tunnel.
10 . A non-transitory computer readable medium having stored thereon instructions for causing a processing circuitry to execute a process, the process comprising:
identifying a plurality of atomic tunnels based on data including packets sent between devices among a plurality of devices, wherein each atomic tunnel is a structure representing communications among the plurality of devices defined with respect to at least three nodes and at least two edges, each node representing a respective device of the plurality of devices, each edge representing a connection between two of the plurality of devices, each atomic tunnel having two hops, wherein each hop is a level of communication in which a packet is sent from one device to another device, and wherein at least one device of the plurality of devices comprises a remotely accessed device accessed by one or more other devices of the plurality of devices using one or more remote system protocols;
identifying a plurality of tunnel constructs, wherein each tunnel construct is a structure including at least one of the plurality of atomic tunnels;
determining a potentially malicious atomic tunnel among the plurality of atomic tunnels by comparing the edges of each of the plurality of atomic tunnels to a plurality of edges of a plurality of previously observed tunnel constructs;
determining a potentially malicious tunnel, wherein the potentially malicious tunnel includes the potentially malicious atomic tunnel, and wherein determining the potentially malicious tunnel comprises:
performing a classification of the plurality of previously observed tunnel constructs to classify each previously observed tunnel construct into a class of a plurality of classes, wherein each class comprises an associated structure;
isomorphically comparing a tunnel structure of the potentially malicious atomic tunnel to the associated structure of each class of the plurality of classes; and
matching the tunnel structure with one of the associated structures to determine the potentially malicious tunnel; and
performing at least one mitigation action based on the determined potentially malicious tunnel.
11 . A system for malicious lateral movement detection, comprising:
a processing circuitry; and
a memory, the memory containing instructions that, when executed by the processing circuitry, configure the system to:
identify a plurality of atomic tunnels based on data including packets sent between devices among a plurality of devices, wherein each atomic tunnel is a structure representing communications among the plurality of devices defined with respect to at least three nodes and at least two edges, each node representing a respective device of the plurality of devices, each edge representing a connection between two of the plurality of devices, each atomic tunnel having two hops, wherein each hop is a level of communication in which a packet is sent from one device to another device, and wherein at least one device of the plurality of devices comprises a remotely accessed device accessed by one or more other devices of the plurality of devices using one or more remote system protocols;
identify a plurality of tunnel constructs, wherein each tunnel construct is a structure including at least one of the plurality of atomic tunnels;
determine a potentially malicious atomic tunnel among the plurality of atomic tunnels by comparing the edges of each of the plurality of atomic tunnels to a plurality of edges of a plurality of previously observed tunnel constructs;
determine a potentially malicious tunnel, wherein the potentially malicious tunnel includes the potentially malicious atomic tunnel, and wherein determining the potentially malicious tunnel comprises:
performing a classification of the plurality of previously observed tunnel constructs to classify each previously observed tunnel construct into a class of a plurality of classes, wherein each class comprises an associated structure;
isomorphically comparing a tunnel structure of the potentially malicious atomic tunnel to the associated structure of each class of the plurality of classes; and
matching the tunnel structure with one of the associated structures to determine the potentially malicious tunnel; and
perform at least one mitigation action based on the determined potentially malicious tunnel.
12 . The system of claim 11 , wherein the potentially malicious atomic tunnel is one of the plurality of atomic tunnels for which at least one of the at least two edges of the potentially malicious atomic tunnel is an outlier edge as compared to the edges of the plurality of previously observed tunnel constructs.
13 . The system of claim 12 , wherein the system is further configured to:
analyze session data for the edges of at least one tunnel construct of the plurality of tunnel constructs in order to determine whether each edge of the analyzed at least one tunnel construct is an outlier edge as compared to the edges of the plurality of previously observed tunnel constructs.
14 . The system of claim 13 , wherein the analyzed session data includes at least one of: time of beginning of session, total session duration, amount of incoming data, and amount of outgoing data.
15 . The system of claim 12 , wherein each edge of the potentially malicious atomic tunnel is either a new edge or an outlier edge as compared to the edges of the plurality of previously observed tunnel constructs, wherein a new edge is an edge that was not included among the plurality of previously observed tunnel constructs.
16 . The system of claim 11 , wherein the system is further configured to:
compare a structure of the potentially malicious tunnel construct to a structure associated with each of a plurality of tunnel classes in order to determine a classification for the potentially malicious tunnel, wherein each of the plurality of tunnel classes is defined based on a respective portion of the plurality of previously observed tunnel constructs;
test a first edge of the potentially malicious tunnel construct in order to determine whether the tested first edge is an outlier edge as compared to a first edge of each of at least one matching class tunnel construct of the plurality of previously observed tunnel constructs, wherein each of the at least one matching class tunnel construct has a same classification as the potentially malicious tunnel construct; and
detect the potentially malicious tunnel as a malicious tunnel when the tested first edge is an outlier edge as compared to the first edge of each of the at least one matching class tunnel construct.
17 . The system of claim 11 , wherein, for each atomic tunnel, a connection time of a first hop of the atomic tunnel includes a connection time of a second hop of the atomic tunnel.
18 . The system of claim 11 , wherein, for each atomic tunnel, an amount of data sent in a first hop of the atomic tunnel includes an amount of data sent in a second hop of the atomic tunnel.
19 . The system of claim 11 , wherein performing the at least one mitigation action includes generating an alert including a risk level, wherein the risk level is determined based on at least one of: a number of devices represented by the nodes of the potentially malicious tunnel, and a number of potentially malicious atomic tunnels included in the potentially malicious tunnel.