Auto-Generating Attack Detection Rules via Network Traffic Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Network-based Attack Detection Systems (ADS) face challenges in timely updating decision rule databases to effectively detect new threats and reducing false positives, as existing systems rely on conventional syntax and manual rule updates, which are inefficient in rapidly evolving computer networks and malware landscapes.

Innovation Solution

The auto-generation of decision rules for attack detection feedback systems, where sensors monitor network traffic, extract metadata, and centrally aggregate information to create new signatures and rules, enabling automated updates and reducing false positives by analyzing communication sessions across multiple protocols and protocols within the OSI model.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual rule updates are used to maintain decision rule databases, then rule accuracy can be controlled, but the system cannot keep up with rapidly evolving threats and the update process is inefficient

Engineering Contradiction:
Improvedetection effectivenessVSAvoidrule update speed
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system automatically generates decision rules by analyzing network traffic data and threat patterns without requiring manual intervention. The ADS autonomously updates its decision rule database by extracting patterns from monitored traffic, enabling self-service maintenance that keeps pace with evolving threats while maintaining detection reliability

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system implements a feedback mechanism where detection results and traffic analysis are continuously fed back into the rule generation process. This closed-loop approach allows the system to learn from actual traffic patterns and threat encounters, automatically refining and updating rules based on real-world performance data

Inventive Principle:
Principle #23Feedback

2Measurement precision

If conventional syntax and manual methods are used for rule creation, then rule precision can be maintained, but the system complexity increases and automation is limited

Engineering Contradiction:
Improvefalse positive rateVSAvoidrule generation automation
Core Design Contradiction:
Measurement precisionVSExtent of automation

Solution Approach 1:

The system replaces manual mechanical rule creation processes with automated data-driven generation. Instead of analysts manually crafting rules based on conventional syntax, the system automatically analyzes network traffic patterns and generates rules algorithmically, substituting human manual work with computational processes that maintain precision while enabling full automation

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system extracts and copies effective rule patterns from analyzed traffic data and threat intelligence sources. By identifying recurring malicious patterns and copying their structural characteristics into new rules, the system maintains detection precision while automating the rule creation process through pattern replication rather than manual construction

Inventive Principle:
Principle #26Copying

Data Source

PatentEP3297248B1System and method for generating rules for attack detection feedback system
Publication Date: 2020.04.29 GRP IB TDS LTD
  • EP3297248B1 patent drawingFigure 1
  • EP3297248B1 patent drawingFigure 2
  • EP3297248B1 patent drawingFigure 3

AI summary

There is provided a method for auto-generation of decision rules for attack detection feedback systems. The method is executed on a server. The method comprises: receiving at least one event from an event database, the event database having been generated from data obtained by at least one sensor; analyzing the at least one event to determine whether the at least one event belongs to a class of malware control center interactions; if the at least one event belongs to the class of malware control center interactions, extracting at least one attribute from the at least one event; generating decision rules using the at least one attribute; and saving the decision rules; saving the decision rules, the decision rules being instrumental in updating what type of further data is obtained by the at least one sensor based on the decision rule.