IP Library › Granted Patent US 10,516,688
Granted Patent B2
US 10,516,688 · App. 15/830,692 · Granted Dec 24, 2019

Ransomware resilient cloud services

Inventors: Gal Tamir (St. Avichayil, IL); Elad Iwanir (Metar, IL)
Assignee: Microsoft Technology Licensing, LLC
H04L63/1425G06F11/1448G06F11/1461G06F11/1464G06F11/3065G06F21/568H04L63/1408G06F11/1469G06F2201/805H04L63/1441
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,516,688
App. No.
15/830,692
Granted
Dec 24, 2019
Kind
B2
Abstract

An anti-ransomware system protects data in cloud storage of a cloud services provider against a ransomware attack. A backup handler is configured to at least one of: selectively retrieve backup data generated by the cloud services provider from the cloud storage; and selectively generate backup data based on the data in the cloud storage and output the backup data to a storage device. A ransomware detector is configured to detect data changes to the data resulting from a ransomware attack. A ransomware remediator communicates with the ransomware detector and the backup handler and is configured to restore the data to a state prior to the ransomware attack based upon the backup data.

Claims (54)

1. An anti-ransomware system for protecting data in cloud storage of a cloud services provider against a ransomware attack, the anti-ransomware system comprising:

one or more processor(s); and

one or more computer-readable hardware storage device(s) having stored thereon computer-executable instructions that are executable by the one or more processor(s) to cause the anti-ransomware system to:

selectively retrieve backup data generated by the cloud services provider from the cloud storage, or, alternatively, selectively generate backup data based on the data in the cloud storage and output the backup data to a storage device;

detect data changes to the data resulting from a ransomware attack;

detect the ransomware attack using at least one of deep learning analysis detection or machine learning, wherein, in one or more instance(s), the ransomware attack is further detected by monitoring one or more change(s) made to honeypot data included in the data, which is stored in the cloud storage;

generate a ransomware alert in response to detecting the ransomware attack;

output the ransomware alert to a queue, wherein, after the queue outputs the ransomware alert, a coordination event occurs to determine how the backup data is to be used in restoring the data, which is subject to the ransomware attack; and

after the queue outputs the ransomware alert, restore the data to a state prior to the ransomware attack based upon the backup data.

2. The system of claim 1 , wherein execution of the computer-executable instructions further causes the system to:

receive change events;

use a plurality of rules to detect the ransomware attack.

3. The system of claim 1 , wherein execution of the computer-executable instructions further causes the system to:

replace the data based upon the backup data in response to the ransomware alert.

4. The system of claim 1 , wherein the cloud storage generates the backup data.

5. The system of claim 1 , wherein detecting the data changes to the data resulting from the ransomware attack is performed using one or more rule(s) used to filter the data in the database.

6. The system of claim 1 , wherein the cloud storage generates the backup data, and wherein detecting the data changes to the data resulting from the ransomware attack is performed using one or more rule(s) used to filter the data in the database.

7. The system of claim 1 , wherein detecting the ransomware attack is performed using the machine learning,

wherein the machine learning uses a support vector machine to distinguish between legitimate changes to the data and malicious changes to the data, and

wherein the support vector machine is configured to support linear classification and nonlinear classification, the nonlinear classification being supported as a result of modifying a kernel function of the system.

8. The system of claim 1 , wherein the system is further configured to use one or more of the following to detect the ransomware attack: a radial basis function (RBF) kernel, a polynomial gaussian RBF kernel, or a sigmoid kernel.

9. An anti-ransomware system for protecting data in cloud storage of a cloud services provider against a ransomware attack, the anti-ransomware system comprising:

a processor;

memory; and

an anti-ransomware detection and remediation application that is instantiated in the memory and that is executed by the processor, the anti-ransomware detection and remediation application being configured to:

selectively retrieve backup data generated by the cloud services provider from the cloud storage, or, alternatively, selectively generate backup data based on the data in the cloud storage and output the backup data to a storage device;

detect data changes to the data resulting from a ransomware attack;

detect the ransomware attack using at least one of deep learning analysis detection or machine learning, wherein, in one or more instance(s), the ransomware attack is further detected by monitoring one or more change(s) made to honeypot data included in the data, which is stored in the cloud storage;

generate a ransomware alert in response to detecting the ransomware attack;

output the ransomware alert to a queue, wherein, after the queue outputs the ransomware alert, a coordination event occurs to determine how the backup data is to be used in restoring the data, which is subject to the ransomware attack; and

after the queue outputs the ransomware alert and in response to the ransomware attack, restore the data to a state prior to the ransomware attack based upon the backup data.

10. The system of claim 9 , wherein the anti-ransomware detection and remediation application is further configured to:

receive events for the data from the cloud storage; and

apply a plurality of rules to the events to detect the ransomware attack.

11. The system of claim 10 , wherein the anti-ransomware detection and remediation application is further configured to detect the ransomware attack based on the events using at least one of the deep learning analysis detection or the machine learning.

12. The system of claim 9 , wherein the anti-ransomware detection and remediation application is further configured to:

receive ransomware alerts; and

replace data based upon the backup data in response to the ransomware alerts.

13. The system of claim 9 , wherein the anti-ransomware detection and remediation application is further configured to schedule backups of the data in response to at least one of an event or a predetermined backup period and to store the backup data in the storage device.

14. The system of claim 9 , wherein the cloud storage generates the backup data.

15. The system of claim 9 , wherein the anti-ransomware detection and remediation application generates the backup data.

16. The system of claim 9 , wherein the cloud storage and the anti-ransomware detection and remediation application generate the backup data.

17. The system of claim 9 , wherein the anti-ransomware detection and remediation application is further configured to detect a ransomware note in the data changes.

18. A method for protecting data in cloud storage of a cloud services provider against a ransomware attack, the method comprising:

selectively retrieving backup data generated by the cloud services provider from the cloud storage, or, alternatively, selectively generating backup data based on the data in the cloud storage and outputting the backup data to a storage device;

detecting data changes to the data resulting from a ransomware attack;

detecting the ransomware attack using at least one of deep learning analysis detection or machine learning, wherein, in one or more instance(s), the ransomware attack is further detected by monitoring one or more change(s) made to honeypot data included in the data which is stored in the cloud storage;

generating a ransomware alert in response to detecting the ransomware attack;

outputting the ransomware alert to a queue, wherein, after the queue outputs the ransomware alert, a coordination event occurs to determine how the backup data is to be used in restoring the data, which is subject to the ransomware attack; and

after the queue outputs the ransomware alert and in response to detecting the ransomware attack, restoring the data to a state prior to the ransomware attack based upon the backup data.

19. The method of claim 18 , further comprising receiving events relating to changes to the data from the cloud storage.

20. The method of claim 19 , further comprising monitoring the events using a plurality of rules to detect the ransomware attack and to generate the ransomware alert.

21. The method of claim 19 , further comprising detecting the ransomware attack and generating the ransomware alert based on the events using at least one of the deep learning analysis detection or the machine learning.

22. The method of claim 18 , wherein the cloud service provider generates the backup data.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Dec 4, 2017
From: TAMIR, GAL; IWANIR, ELAD
To: MICROSOFT TECHNOLOGY LICENSING, LLC
Reel/Frame 044289/0576 →
Continuity (2)
Provisional Application 62449387 · Jan 23, 2017
Related Publication 20180212987A1 · Jul 26, 2018
Cited By (15)
US 12,197,398 US 12,204,657 US 12,229,261 US 12,242,455 US 12,248,434 US 12,248,435 US 12,248,566 US 12,367,108 US 12,379,999 US 12,411,962 US 12,457,214 US 12,517,874 US 12,554,849 US 12,561,428 US 12,627,681