System and method for predicting impending cyber security events using multi channel behavioral analysis in a distributed computing environment

• US 9,602,530 B2
• Filed: 03/30/2015
• Issued: 03/21/2017
• Est. Priority Date: 03/28/2014
• Status: Expired due to Fees

• Current Assignee: Zitovault, Inc.
• Original Assignee: Zitovault, Inc.
• Inventors: Ellis, Alonzo, McELwee, Tim
• Primary Examiner(s): Waliullah, Mohammed
• Application Number:
  US14/672,869
• Publication Number:
  US 20160028758A1
• Time in Patent Office: 722 Days
• Field of Search: 726 22- 25
• US Class Current: 1/1
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06F 21/52 during program execution, e.g. stack integrity ; Preventing unwanted data erasure; Buffer overflow
• CPC Class Codes: G06F 21/554 involving event detection and direct action
• CPC Class Codes: G06F 21/577 Assessing vulnerabilities and evaluating computer system security
• CPC Class Codes: G06Q 30/018 Certifying business or products
• CPC Class Codes: G06Q 30/0185 Product, service or business identity fraud
• CPC Class Codes: H04L 63/0227 Filtering policies mail message filtering H04L51/212
• CPC Class Codes: H04L 63/14 for detecting or protecting against malicious traffic
• 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
• CPC Class Codes: H04L 63/1433 Vulnerability analysis
• CPC Class Codes: H04L 63/1441 Countermeasures against malicious traffic countermeasures against attacks on cryptographic mechanisms H04L9/002
• CPC Class Codes: H04L 63/20 for managing network security; network security policies in general filtering policies H04L63/0227

Abstract

Multi channel distributed behavioral analysis architecture provides a software solution to the major operational challenges faced with providing an early warning system for impending cyber security events. Most cyber security events are premeditated. However, many current cyber security defense technologies only address the real-time detection of a software vulnerability, the presence of malware (known or unknown “zero day”), anomalies from pre-established data points, or the signature of an active security event. The system and method of the multi channel distributed behavioral analysis architecture introduces a technique which provides the data collection, assessment, and alerting ability prior to the occurrence of an event based on threat actor behavior.

Claims

1. A method for predicting the likelihood future security threats in distributed computing environment comprising the steps of:

a main server entity communicating with a plurality of decision engines and a plurality of correlation engines;

a communication engine communicating with the main server;

the communication engine communicating with one or more third parties, wherein the one or more third parties comprises one or more data points, wherein the one or more data points comprise a plurality of user data and event data;

a deep learning engine receiving outputs from multiple simulation runs using sequential game theory against the one or more data points, wherein the event data is selected from the group consisting of temporal, geographic, social, financial and linguistic data;

the deep learning engine predicting a first occurrence of a security event, wherein predicting the first occurrence of a security event is based on the receiving the outputs from the multiple simulation runs;

building a plurality of semantic graphs based on the communicating with correlation engines;

a plurality of distributed networked agents collecting event and attribute data for the main server entity, the plurality of correlation server engines, and the plurality of decision engines, wherein the plurality of distributed networked agents are maintained on local servers; and

a defined protocol initiating and maintaining secure communication between the main server, the plurality of distributed networked agents, the plurality of correlation engines, the plurality of decision engines and the communication engine;

forecasting an arrival time of the impending security threat by incorporating temporal data in a prediction process, wherein the prediction process comprises;

discovering said agent servers;

determining an available processing bandwidth of the main server, agents, decision engines, and correlation engines;

registering said main server and available agent server with registration entity;

correlating event and attribute data from unstructured data sources;

collecting log data;

normalizing and tokenize the log data;

generating semantic graphs of the log and the attribute data; and

deciding on the likelihood of an event using the sequential game theory.

2. The method according to claim 1, wherein the step of deciding on the likelihood of an event using the sequential game theory, further comprises:

discovering attempts to hide an observation of events by staging other events;

discovering attempts to hide the observation of events by utilizing a plurality of communications channels to execute an attack; and

making predictions in the event of partial data loss.

3. The method according to claim 1, wherein the method for distributing workload further comprises:

breaking data analysis work across an virtually unlimited amount cpu cores in order to scale data processing to handle an unlimited amount or log and attribute data; and

shifting workload in the event of a catastrophic loss of processing agents with minimal loss of fidelity, wherein auto provision of processing agents as bandwidth, log and processing demands increase or decrease.

4. The method according to claim 2, wherein the step of collecting log data further comprises:

self healing databases while still processing transactions due to a security attack, a loss/addition of one or more virtual machine cores, a memory/disk buffer overwrite, a major shift in workload, or a hardware failure.