Exploring Security Solutions and Vulnerabilities for Embedded Non-Volatile Memories
Zakia Tamanna Tisha, Jeremy Muldavin, Ujjwal Guin · 2024
Non-volatile memories (NVMs) have gained prominence in the modern semiconductor industry as they strive to address the requirements of high-performance computing. NVMs have become appealing components in secure systems due to their limited area-energy-runtime budgets and extended data retention capabilities. However, the data retention capability of NVMs, which persists even after power is switched off, also makes them vulnerable to malicious attacks, raising security concerns about their applications. Despite the considerable number of publications on the security applications and vulnerabilities of NVMs separately, few studies have combined these aspects in recent years. During this period, significant developments have occurred in the NVM landscape, including the creation of new hardware security primitives based on NVMs and the emergence of new attacks against them. This paper aims to fill the research gap by providing a comprehensive review of both the security applications and vulnerabilities of embedded NVMs, serving as a valuable resource for researchers new to this area. The paper discusses research trends in NVM-based security primitives such as Physically Unclonable Functions (PUFs) and True Random Number Generators (TRNGs), as well as logic obfuscation techniques. It also explores the risks and common attacks associated with embedded NVMs and concludes with suggestions for future research directions in the security aspects of this domain.