This work presents a power side-channel analysis (SCA) of a lightweight cryptography (LWC) algorithm, XOODYAK, implemented on an FPGA. First, we perform generic leakage detection tests for two phases of authenticated encryption with associated data (AEAD) mode, namely INITIALIZE, and ABSORB. Second, we develop novel hypothetical attack models for correlation power analysis (CPA) and demonstrate a success rate (SR) of 92%/82% and minimum-traces-to-disclosure (MTD)=13K/38K on the INITIALIZE/ABSORB phases, respectively. Third, we evaluate ABSORB against Profiled SCA using convolutional neural network (CNN), and achieve SR=96%/64% and MTD=2K/16K on the test set for the same/different keys used for training, respectively. Finally, we suggest low-overhead countermeasures to protect against these SCA attacks.

In this work, we address the challenges of side channel analysis (SCA) and fault injection (FI) in the pre-silicon design stage, particularly for cryptographic implementations. We show that the industry perspective uncovers different challenges than those usually tackled by academia, emphasizing the need for practical and scalable solutions. We present (Side Channel Attack Testbench Emulator) SCATE, a framework for detecting and mitigating SCA vulnerabilities, leveraging industry-standard electronic design automation (EDA) software for efficient power estimation and FI simulations. Moreover, we demonstrate a significant performance increase in power simulation tools by optimizing EDA algorithms for SCA applications, resulting in faster power trace generation. Additionally, we discuss the application of formal property verification (FPV) for verifying fault injection countermeasures and present a case study on an AES S-box. These approaches enable non-security-expert designers to evaluate SCA and FI resistant designs with significant time savings. The vulnerabilities uncovered by these approaches highlight the importance of pre-silicon analysis at each design stage to produce secure application-specific integrated circuits (ASICs).