FlagThis

Abstract

Most corporate workplace environments enforce policies and technical controls that limit the storage of sensitive data on client endpoints. Consequently, ransomware operators have evolved variants that expand their attack surface from local systems to network drives and shared storage resources. As traditional endpoint detection mechanisms focus primarily on local system behaviour, a compromised client can impact remote fle servers, such as by encrypting shared data, without directly triggering behavioural changes on the servers themselves. In this paper, we propose a hybrid detection framework for detecting crypto-ransomware intrusion within integrated fle server and client environments. The framework is based on a new technique referred to as Region of Interest (RoI) to analyse network trafc and extract Indicators of Compromise (IoCs). The IoC repository serves as an additional ruleset to enhance existing security tools such as EDRs and IDSs, while RoI-derived features are used to train an ML model to detect highly evasive variants. This study incorporates a broader set of ransomware families and carefully selected benign behaviours based on domain expertise, ensuring coverage of common user actions that could interfere with ransomware detection. Beyond IoCs, which operate in a signature-based manner, our machine learning module achieves a detection precision of 99.64%, with a 0% false negative rate (FNR) and a minimal false positive rate (FPR). Furthermore, the proposed method enables early detection, identifying ransomware intrusions before signifcant damage occurs, achieving an accuracy of 99.44%.