A self-healing redundancy scheme for mission/safety-critical applications

• Main Authors: Balasubramanian, Padmanabhan, Maskell, Douglas Leslie
• Other Authors: School of Computer Science and Engineering
• Format: Article
• Language: English
• Published: 2019
• Subjects: Fault Tolerance; DRNTU::Engineering::Computer science and engineering; Redundancy
• Online Access: https://hdl.handle.net/10356/103389; http://hdl.handle.net/10220/47290
• Institution: Nanyang Technological University

In the nanoelectronics era, multiple faults or failures in circuits and systems deployed in mission- and safety-critical applications, such as space, aerospace, nuclear etc., are known to occur. To withstand these, higher order redundancy is suggested to be used selectively in the sensitive portions of a circuit or system. In this context, the distributed minority and majority voting based redundancy (DMMR) scheme was proposed as an alternative to the N-modular redundancy (NMR) scheme for the efficient implementation of higher order redundancy. However, the DMMR scheme is not self-healing. In this paper, we present a new self-healing redundancy (SHR) scheme that can inherently correct its internal faults or failures without any external intervention, which makes it ideal for mission/safety-critical applications. To achieve the same degree of fault tolerance, the SHR scheme requires fewer function blocks than the NMR and DMMR schemes. We present the architectures of the proposed SHR scheme, discuss the system reliability, and provide the design metrics estimated for example SHR systems alongside the corresponding NMR and DMMR systems using a 32/28-nm CMOS technology. From the perspectives of fault tolerance, self-healing capability, and optimizations in the design metrics, the SHR scheme is preferable to the NMR and DMMR schemes.