Malware Detection and Encryption: A Meta-Survey With a Unified Analytical Framework

This paper presents a meta-survey of the current state of the art in malware detection and
the use of cryptographic evasion and obfuscation techniques in malicious software. Based on a corpus of
over 70 selected studies, it synthesises heterogeneous survey findings via a unified analytical framework.
Then, as the topic of crypto-malware is largely absent from the academic survey corpus, a narrative review
of those mechanisms is provided, drawing on primary technical literature and industry threat intelligence.
The paper identifies four principal findings: malware detection is increasingly framed as a representationaware learning pipeline; robustness must be treated as a first-order qualifier of any state-of-the-art claim;
the cryptographic and steganographic dimension of malware is underrepresented in the survey literature
relative to its operational relevance; and persistent white spaces remain in the literature, particularly for IoT
graph-based detection, Android image-based detection, and crypto/stego-aware detection pipelines