A Fault Analysis and Classifier Framework for Reliability-Aware SRAM-Based FPGA Systems

Abstract

This paper presents a new framework for the analysis of SRAM-based FPGA systems with respect to their dependability properties against single, multiple and cumulative upsets errors. The aim is to offer an environment for performing fault classification and error propagation analyses for designed featuring fault detection or tolerance techniques against soft errors, where the focus is not only the overall achieved fault coverage, but an understanding of the fault/error relation inside the internal elements of the system. We propose a fault analyzer/classifier laying on top of a classical fault injection engine, used to monitor the evolution of the system after a fault as occurred, with respect to the applied reliability-oriented design technique. The paper introduces the framework and reports some experimental results of its application to a case study, to highlight the benefits of the proposed solution.