Detecting cryptographic ransomware by examining file system activity
Year: 2019
Type: dissertation
Abstract: Cryptographic ransomware, which locks a victim’s files and demands payment to reestablish access, is one of the most dangerous and popular cyber crimes of today as it gives attackers a golden opportunity to extort money. Although many different approaches are presented to detect and prevent this troublesome malware, recent research suggests that none of these approaches are flawless and they can be bypassed. \nIn this thesis, we propose CryptoCop, a protection system that stops a ransomware attack in the early stages. The defense mechanism limits the applications’ capability of executing file write functions, which is excessively performed by a typical ransomware. We define an adaptive threshold mechanism for file write requests of each running process, which facilitates benign file system operations while terminating the malicious activity. The results of experiments show that CryptoCop is able to stop 706 out of 736 (⇠96%) ransomware samples with minimal loss of files – less than 5 – and a negligible performance overhead.
Author: Akay, Meltem
Language: English
Subfield: Software
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W3123696521
