Ransomware-Resilient File Backup System with Anomaly Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing security solutions are inadequate in detecting and managing ransomware attacks, particularly cryptoviral extortion, which can lead to complete data loss due to the large encryption keys involved, making it difficult to recover files without the encryption key.

Innovation Solution

The system detects anomalies during file backups that may indicate ransomware, isolates potentially encrypted backup copies, and adjusts the backup policy to prevent overwriting with existing copies, allowing for safe recovery by discarding affected backup copies and reverting to unencrypted versions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If backup copies are created in a separate location to protect against ransomware, then data recovery capability is improved, but backup policy complexity increases

Engineering Contradiction:
Improvedata recovery capabilityVSAvoidbackup policy complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The backup system segments backup copies into different storage locations based on detected ransomware threats. When ransomware is detected, the system creates isolated backup copies in a separate location rather than following the normal backup rotation policy, thereby protecting recovery capability while managing complexity through conditional segmentation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The backup policy dynamically adjusts based on real-time ransomware detection. The system transitions from a static backup policy to a dynamic one that modifies backup behavior (creating separate copies) only when threats are detected, balancing reliability improvement with minimal policy complexity increase.

Inventive Principle:
Principle #15Dynamics

2Measurement precision

If anomaly detection is performed during file backup to identify ransomware, then detection accuracy is improved, but processing time increases

Engineering Contradiction:
Improveransomware detection accuracyVSAvoidbackup processing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The system performs partial anomaly detection during backup operations by monitoring specific indicators (file entropy, modification patterns) rather than conducting exhaustive analysis on every file. This approach maintains adequate detection accuracy while minimizing the time penalty during backup processing.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

When ransomware is detected through anomaly detection, the system skips normal backup operations for affected files and immediately isolates them. This rushing through the detection and response process minimizes overall processing time by avoiding lengthy analysis once a threat is identified.

Inventive Principle:
Principle #21Skipping (Rushing through)

Data Source

PatentUS10742665B2Systems and methods for modifying file backups in response to detecting potential ransomware
Publication Date: 2020.08.11 GEN DIGITAL INC
  • US10742665B2 patent drawing
  • US10742665B2 patent drawing
  • US10742665B2 patent drawing

AI summary

A computer-implemented method for modifying file backups in response to detecting potential ransomware may include (1) detecting, during a file backup process, an anomaly that is potentially indicative of ransomware on a computing device, (2) in response to detecting the anomaly that is potentially indicative of ransomware, storing a backup copy in a separate location from other backup copies, (3) confirming that the anomaly is indicative of ransomware on the computing device, (4) adjusting a backup policy in response to confirming that the anomaly is indicative of ransomware on the computing device, and (5) managing the backup copy based on the adjusted backup policy. Various other methods, systems, and computer-readable media are also disclosed.