Memory Analysis for Malware Detection: A Comprehensive Survey Using the OSCAR Methodology

Abstract

The steady growth of malware over the years has now sharply escalated, with a 30% surge in global cyberattacks in 2024. This rise demands advanced detection, as traditional methods often miss sophisticated or fileless malware. Memory analysis detects traces left by any malware in volatile memory, revealing runtime behaviors, privilege escalation attempts, and active processes. An examination of prior research shows that existing surveys on memory analysis have significant gaps, as none provide a comprehensive overview of the field. To address these gaps, this survey systematically proposes key research questions and addresses them using the OSCAR (Obtain, Strategize, Collect, Analyze, Report) methodology. Memory acquisition techniques and tools have been discussed with the most diverse taxonomy provided to the best of our knowledge. Furthermore, forensic methods, tools, and studies are categorized into four distinct approaches, with a comprehensive taxonomy at the end. We also evaluated and ranked memory dump datasets using our proposed scoring system. Finally, the survey covers malware detection methods, examining both machine learning and traditional approaches and their accuracy, benefits, drawbacks, and challenges. This survey aims to provide a comprehensive and up-to-date overview of the field of memory analysis, with a focus on detecting malicious activities.