A comprehensive survey on hardware-assisted malware analysis and primitive techniques

Abstract

Malware poses an extremely dangerous threat to the digital world, significantly impacting various domains such as smart cities, intelligent transportation, wireless sensor networks (WSNs), and the Internet of Vehicles (IoV). In the Internet of Things (IoT) context, which includes various applications, security threats, and vulnerabilities are high, especially for resource-constrained single-board devices. One of the critical challenges is the presence of unauthorised access to IoT devices that attempt to replicate authorized devices by mimicking their hardware and software specifications. This review paper provides a comprehensive overview of the security issues related to various hardware-level attacks, memory attacks, hardware Trojans, and their countermeasures in different domains. In addition, the paper addresses the study of hardware malware and explores their detection techniques for hardware-based malware, including side-channel and chip-based attacks using machine learning. Using primitive hardware techniques, such as Physically Unclonable Functions (PUFs), to generate the unique key for the device by preventing unauthorised access to the hardware during authentication for different domains is thoroughly investigated with different security features. This review analyzes the different types of PUFs, their characteristics, and their use for various security tasks, including device authentication and the key generation process. It critically evaluates the strengths and limitations of the PUF-based approach and identifies potential research directions to overcome the existing challenges.