Enhanced cybersecurity using an automated planning service
A system and methods for enhanced cybersecurity using an automated planning service is provided, comprising an automated planning service that receives a network analysis job, constructs a simulation using known network and threat information, assigns individual actions to worker nodes for processing, analyzes the results and records new threat mitigation plans as they are produced to expedite future operations, and produces an action plan as output.
1 . A system for enhanced cybersecurity using an automated planning service, comprising:
an automated planning service comprising a memory, a processor, and a plurality of programming instructions stored in the memory thereof and operable on the processor thereof, wherein the programmable instructions, when operating on the processor, cause the processor to:
operate a plurality of master computing nodes, each master computing node in turn operating a plurality of worker computing nodes;
receive a network analysis job at a first master computing node;
receive, from a distributed computational graph, a plurality of security parameters and threat definitions, wherein the plurality of security parameters and threat definitions are based, at least in part, on streaming data processed by the distributed computational graph;
identify a security threat based on the network analysis job and the plurality of security parameters and threat definitions;
determine a desired network end state, wherein the desired network end state is based on the network analysis job and the plurality of security parameters and threat definitions;
assign, using the first master computing node, a plurality of actions to a plurality of worker computing nodes, wherein each of the plurality of actions is based on the identified security threat, the network analysis job, and the desired network end state, wherein each of the plurality of worker computing nodes is assigned exactly one of the plurality of actions at any given time during operation;
receive, at the first master computing node, results of the plurality of actions from the plurality of worker computing nodes as the actions are completed;
analyze, using the first master computing node, the received results to determine which of the actions have brought a current network state closer to the desired network end state;
iteratively instruct, using the first master computing node, at least one of the worker computing nodes to perform at least one additional action, wherein the at least one additional action performed by the at least one worker computing node is based on the analysis of the received results;
when the desired network end state is reached, generate a threat mitigation plan for the identified security threat comprising a record of the actions that led to the desired network end state and the order in which said actions were completed; and
provide the analyzed results and the threat mitigation plan as output.
2 . The system of claim 1 , wherein at least a portion of the plurality of security parameters and threat definitions is retrieved from a data storage.
3 . The system of claim 1 , wherein at least a portion of the plurality of security parameters and threat definitions is received via a representational state transfer (RESTful) application programming interface (API).
4 . The system of claim 1 , wherein the streaming data processed by the distributed computational graph originates from a plurality of sensors and electronic infrastructure monitors associated with the system.
5 . The system of claim 1 , wherein the plurality of security parameters and threat definitions are updated in near real-time based on the streaming data.
6 . The system of claim 1 , wherein the network analysis job is updated in near real-time based on the streaming data.
7 . The system of claim 1 , wherein each worker computing node operates independently from each other worker computing node in an asynchronous “eventual agreement” model.
8 . The system of claim 1 , wherein the distributed computational graph comprises a plurality of transformation pipelines arranged as a directed graph of transformation nodes and messages.
9 . The system of claim 8 , wherein the distributed computational graph is configured via representational state transfer (RESTful) application programming interface (API) endpoints to start, stop, and update pipelines without halting other pipelines.
10 . The system of claim 8 , wherein one or more of the plurality of transformation pipelines of the distributed computational graph is updated in near-real-time based on the streaming data.
11 . A method for enhanced cybersecurity using an automated planning service, comprising the steps of:
operating, at an automated planning service, a plurality of master computing nodes, each master computing node in turn operating a plurality of worker computing nodes;
receiving a network analysis job at a first master computing node;
receiving, from a distributed computational graph, a plurality of security parameters and threat definitions, wherein the plurality of security parameters and threat definitions are based, at least in part, on streaming data processed by the distributed computational graph;
identifying a security threat based on the network analysis job and the plurality of security parameters and threat definitions;
determining a desired network end state, wherein the desired network end state is based on the network analysis job and the plurality of security parameters and threat definitions;
assigning, using the first master computing node, a plurality of actions to a plurality of worker computing nodes, wherein each of the plurality of actions is based on the identified security threat, the network analysis job, and the desired network end state, wherein each of the plurality of worker computing nodes is assigned exactly one of the plurality of actions at any given time during operation;
receiving, at the first master computing node, results of the plurality of actions from the plurality of worker computing nodes as the actions are completed;
analyzing, using the first master computing node, the received results to determine which of the actions have brought a current network state closer to the desired network end state;
iteratively instructing, using the first master computing node, at least one of the worker computing nodes to perform at least one additional action, wherein the at least one additional action performed by the at least one worker computing node is based on the analysis of the received results;
when the desired network end state is reached, generating a threat mitigation plan for the identified security threat comprising a record of the actions that led to the desired network end state and the order in which said actions were completed; and
providing the analyzed results and the threat mitigation plan as output.
12 . The method of claim 11 , wherein at least a portion of the plurality of security parameters and threat definitions are is retrieved from a data storage.
13 . The method of claim 11 , wherein at least a portion of the plurality of security parameters and threat definitions is received via a representational state transfer (RESTful) application programming interface (API).
14 . The method of claim 11 , wherein the streaming data processed by the distributed computational graph originates from a plurality of sensors and electronic infrastructure monitors associated with the system.
15 . The method of claim 11 , wherein the plurality of security parameters and threat definitions are updated in near real-time based on the streaming data.
16 . The method of claim 11 , wherein the network analysis job is updated in near real-time based on the streaming data.
17 . The method of claim 11 , wherein each worker computing node operates independently from each other worker computing node in an asynchronous “eventual agreement” model.
18 . The method of claim 11 , wherein the distributed computational graph comprises a plurality of transformation pipelines arranged as a directed graph of transformation nodes and messages.
19 . The method of claim 18 , wherein the distributed computational graph is configured via representational state transfer (RESTful) application programming interface (API) endpoints to start, stop, and update pipelines without halting other pipelines.
20 . The method of claim 18 , wherein one or more of the plurality of transformation pipelines of the distributed computational graph is updated in near-real-time based on the streaming data.