A Survey of Fault-Injection Methodologies for Soft Error Rate Modeling in Systems-on-Chips

Abstract

The development of process technology has increased system performance, but the system failure probability has also significantly increased. It is important to consider the system reliability in addition to the cost, performance, and power consumption. In this paper, we describe the types of faults that occur in a system and where these faults originate. Then, fault-injection techniques, which are used to characterize the fault rate of a system-on-chip (SoC), are investigated to provide a guideline to SoC designers for the realization of resilient SoCs.

Fault Injection with Multiple Fault Patterns for Experimental Evaluation of Demand-Controlled Ventilation and Heating Systems

“…FI brings high controllability and observability in a simulation environment. FI techniques are classified into four categories: (1) physical FI, including hardware-based FI and software-based FI methods; (2) simulation-based FI methods; (3) emulation-based FI methods; and (4) hybrid FI methods [ 14 , 15 ]. Arlat et al [ 16 ] have introduced an FI methodology for two main goals: validation and design aid.…”

Realistic Simulation of Sensor/Actuator Faults for a Dependability Evaluation of Demand-Controlled Ventilation and Heating Systems

“…In a simulation framework, faults can be injected using a set of input patterns via an automated FI code or FI dashboard in hardware or software. Several surveys studied FI methodologies [3,[6][7][8][9]. Briefly, FI techniques can be categorized into four methodologies: (1) physical fault injections, including hardware-based fault injection (HaFI) and software-based fault injection (SoFI) methods; (2) simulation-based fault injection (SiFI) methods; (3) emulation-based fault injection (EmFI) methods; and (4) hybrid fault injection (HyFI) methods [4,6,8,10].…”

“…Several surveys studied FI methodologies [3,[6][7][8][9]. Briefly, FI techniques can be categorized into four methodologies: (1) physical fault injections, including hardware-based fault injection (HaFI) and software-based fault injection (SoFI) methods; (2) simulation-based fault injection (SiFI) methods; (3) emulation-based fault injection (EmFI) methods; and (4) hybrid fault injection (HyFI) methods [4,6,8,10]. The advantages and disadvantages of each method were systematically discussed in [7].…”

“…As explained in the related work section, many faults injection models and frameworks in the literature are based on unrealistic sensor and actuator fault models, assuming restricted fault types or deterministic behaviors under faults. Additionally, Gaussian distributions and white noise with a uniform distribution are commonly used to describe the uncertainties of measurements [8]. This paper provides an extended version of the simulation-based fault injection framework based on previous works and the authors in this paper conclude from the table that to provide a realistic simulation of, an SiFI framework for HVAC systems, the FI model must include many fault types, fault persistence, fault durations, fault interarrival times, and the probabilistic variations of observations in the presence of different faults.…”

“…Different fault types can also be described by random variables with corresponding probability distributions. This paper introduces sensor models based on the technical ISO/IEC guidelines [8] for uncertainty in measurements (GUM) for both regular behavior and fault scenarios.…”