Isolation design flow effectiveness evaluation methodology for zynq SoCs

Static Random-Access Memory (SRAM)-based Field Programmable Gate Arrays (FPGAs) are increasingly being used in many application domains due to their higher logic density and reconfiguration capabilities. However, with state-of-the-art FPGAs being manufactured in the latest technology nodes, reliability is becoming an important issue, particularly for safety-critical avionics, automotive, aerospace, industrial robotics, medical, and financial systems. Therefore, fault tolerant system design methodologies have become essential in the aforementioned application domains. The Isolation Design Flow (IDF) is one such design methodology that has promising prospects due to its ability to isolate logic design modules at the physical level for fault containment purposes. This paper proposes a methodology to evaluate the effectiveness of the IDF. To do so, reverse engineering is used to enable fault injection on the IDF designs with minimal changes in the bit-stream. This reduces the time needed to inject a fault significantly thus accelerating the evaluation process. Then this methodology is applied to a case study of a single-chip cryptography application on a ZynQ SoC. Specifically, an Advanced Encryption Standard (AES) Duplication With Comparison (DWC) design is physically isolated with IDF and subsequently subjected to frame-level Fault Injection (FI) in the configuration memory.