基于忆阻器的物理不可克隆函数: 设计、 防护和应用 进展

With the rapid proliferation of Internet of Things (IoT) devices, hardware security is facing significant challenges. Physical unclonable function (PUF), a security technology based on physical properties, leverage the inherent random microvariations that occur during the manufacturing process to generate unique device identifiers, thereby providing robust authentication. Nevertheless, conventional CMOS-based PUFs struggle to meet the stringent requirements of modern applications, owing to limitations in power efficiency and performance. Resistive random-access memory (RRAM), characterized by its high density, low power consumption, and non-volatility, has emerged as a promising candidate for lightweight PUF technology. This work systematically reviews the research progress on RRAM PUFs. It begins by outlining the fundamental principles of PUFs and RRAM, followed by an exploration of RRAM PUF design strategies from two perspectives: material innovation and architectural optimization. These strategies include the development of novel memristive materials, RRAM-specific architectural designs, and integration of RRAM into CMOS-compatible frameworks. To enhance security, defense mechanisms such as reconfiguration, XOR-based techniques, and permutation strategies are thoroughly discussed. In addition, this work highlights the deployment of RRAM PUFs in various real-world applications. Finally, future research directions are proposed, including the development of multidimensional entropy sources, reconfigurable architectures, and collaborative hardware-software optimization, aiming to advance both academic exploration and industrial adoption of RRAM PUF technologies.