| Sumario: | The present study offers an exhaustive systematic review of various ransomware detection techniques, with a specific focus on network traffic analysis and anomaly detection algorithms. In accordance with the PRISMA protocol, this study analyzes fifteen research papers published between 2021 and 2025. The research papers are collected from prominent scientific database sources, including IEEE Xplore, Scopus, ScienceDirect, and Google Scholar. The corpus of research articles can be broadly classified into four main categories of ransomware detection techniques: supervised machine learning approaches, unsupervised anomaly detection techniques, deep learning approaches, and hybrid approaches for ransomware detection.A comparative analysis of the research articles reveals that the deep learning approaches have been able to achieve higher detection accuracies of more than 99%, while the unsupervised anomaly detection approaches have been able to demonstrate higher adaptability in the detection of ransomware. However, the area of ransomware detection has several research gaps that need to be addressed: the lack of standardized data sets for the experiments, the processing of encrypted network traffic, and the occurrence of false positives in the detection process. The research demonstrates the need to incorporate hybrid and behavioral detection techniques to strengthen the robustness of cybersecurity systems against new ransomware attacks.
|
|---|