Recent Advancements in Microarchitectural Security: Review of Machine Learning Countermeasures

Recent microarchitectural security countermeasures by employing applications' low-level features collected from Hardware Performance Counters (HPCs) registers have emerged as a promising solution to address the inefficiency of traditional software-based methods. Furthermore, recent studies have shown that malicious activities at the hardware level ranging from application-based malware to microarchitecture-based Side-Channel Attacks (SCAs) can be accurately distinguished from normal traces using Machine Learning (ML) algorithms. Such ML-based countermeasures reduce the latency of attack detection process as well as the hardware and resource utilization overheads. This paper provides an in-depth analysis of recent advancements in machine learning countermeasures for microarchitectural security to detect malicious software and emerging side-channel attacks.