A Computational Model for Ransomware Detection Using Cross-Domain Entropy Signatures

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Mannon, Michael, Statham, Evan, Featherstone, Quentin, Arkwright, Sebastian, Fenwick, Clive, Willoughby, Gareth
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866915213573357568
author Mannon, Michael
Statham, Evan
Featherstone, Quentin
Arkwright, Sebastian
Fenwick, Clive
Willoughby, Gareth
author_facet Mannon, Michael
Statham, Evan
Featherstone, Quentin
Arkwright, Sebastian
Fenwick, Clive
Willoughby, Gareth
contents Detecting encryption-driven cyber threats remains a large challenge due to the evolving techniques employed to evade traditional detection mechanisms. An entropy-based computational framework was introduced to analyze multi-domain system variations, enabling the identification of malicious encryption behaviors through entropy deviations. By integrating entropy patterns across file operations, memory allocations, and network transmissions, a detection methodology was developed to differentiate between benign and ransomware-induced entropy shifts. A mathematical model was formulated to quantify entropy dynamics, incorporating time-dependent variations and weighted domain contributions to enhance anomaly detection. Experimental evaluations demonstrated that the proposed approach achieved high accuracy across diverse ransomware families while maintaining low false positive rates. Computational efficiency analysis indicated minimal processing overhead, suggesting feasibility for real-time implementation in security-sensitive environments. The study highlighted entropy fluctuations as a useful indicator for identifying malicious encryption processes, reinforcing entropy-driven methodologies as a viable component of cybersecurity strategies.
format Preprint
id arxiv_https___arxiv_org_abs_2502_10711
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Computational Model for Ransomware Detection Using Cross-Domain Entropy Signatures
Mannon, Michael
Statham, Evan
Featherstone, Quentin
Arkwright, Sebastian
Fenwick, Clive
Willoughby, Gareth
Cryptography and Security
Detecting encryption-driven cyber threats remains a large challenge due to the evolving techniques employed to evade traditional detection mechanisms. An entropy-based computational framework was introduced to analyze multi-domain system variations, enabling the identification of malicious encryption behaviors through entropy deviations. By integrating entropy patterns across file operations, memory allocations, and network transmissions, a detection methodology was developed to differentiate between benign and ransomware-induced entropy shifts. A mathematical model was formulated to quantify entropy dynamics, incorporating time-dependent variations and weighted domain contributions to enhance anomaly detection. Experimental evaluations demonstrated that the proposed approach achieved high accuracy across diverse ransomware families while maintaining low false positive rates. Computational efficiency analysis indicated minimal processing overhead, suggesting feasibility for real-time implementation in security-sensitive environments. The study highlighted entropy fluctuations as a useful indicator for identifying malicious encryption processes, reinforcing entropy-driven methodologies as a viable component of cybersecurity strategies.
title A Computational Model for Ransomware Detection Using Cross-Domain Entropy Signatures
topic Cryptography and Security
url https://arxiv.org/abs/2502.10711