Fault-tolerant Solution Research Articles

Effective Fault Effects Evaluation for Permanent Faults in GPUs executing DNNs

Deep Neural Networks (DNNs) have permeated multiple applications, including cutting-edge safety-critical domains, which require relevant computational power, often provided by Graphic Processing Units (GPUs). GPUs are manufactured with advanced semiconductor technologies that can be affected by faults during the operational phase (e.g., due to wear-out, aging, or environmental harshness), whose effects possibly reach the DNN outputs, in some cases leading to catastrophic consequences. Hence, hardware-aware reliability assessments of DNNs are crucial to be considered in the context of safety-critical systems (following regulations/standards of specific application domains). Application-level fault injection (FI) techniques (i.e., DNN parameter corruption) are often adopted for the reliability evaluation of DNNs; unfortunately, these approaches hardly represent fault effects from GPU hardware. This work proposes an FI strategy based on Hardware-Injection-Through-Program-Transformation (HITPT) to mimic the effect of permanent faults (PFs) at the GPU instruction level, enabling effective assessment of PFs on DNN’s reliability. Our approach provides a good trade-off between the fault effect evaluation’s accuracy and the required computational time. Using the proposed approach, for the first time, we systematically assessed the effects of PF in GPUs executing some DNN sample cases. The results indicate that the faults injected closer to the hardware, using our evaluation strategy, can produce a higher accuracy degradation than the evaluations performed by the typical application-level FI that modify only the DNN parameters. Furthermore, the proposed FI methodology provides insightful results to identify the most suitable fault-tolerance solutions (e.g., selective hardening or design diversity) for their application at thread levels inside GPU’s kernels.

Gorilla Troop Optimizer-Driven Fault-Tolerant Scheduling for Cloud-Based Business Workflows

The study proposes a GTO-FTASS (Gorilla Troop Optimizer-Based Fault Tolerant Aware Scheduling Scheme) for improving the reliability and performance in the cloud computing context. Cloud systems are more likely to fail due to the architecture of these layers and dependence on both the hardware and software, therefore require more sophisticated fault-tolerant solutions. The preliminary to this work is the design of an adaptive GTO-FTASS with a fitness function based on two constraints: Expected Time of Completion (ETC) and Failure probability that were derived from the gorilla value system. The approach provides resource utilization and task planning with the provision of fault recovery hence reducing exposure to time loss and operational vulnerability. MGS outperforms several state-of-the-art models, such as MTCT, MAXMIN, ACO, NSGA-II, and DCLCA in terms of makes pan, failure ratio and failure slowdown. Finally, the applicability of experimental validation with various situations and fluctuating intensities demonstrates the scalability of the model and its stability under pressure, decreased failure rates and increased effectiveness of performed tasks. Through the approaches to latency, resource, and error correction, GTO-FTASS is an investment that stewards have to make to cut costs and achieve high performance on clouds. The framework also provides competitive benefit and robustness for cloud enterprising in fluctuating and crucial strategic applications.

Research for an enhanced fault-tolerant solution against the current sensor fault types in induction motor drives

Introduction. Recently, three-phase induction motor drives have been widely used in industrial applications; however, the feedback signal failures of current sensors can seriously degrade the operation performance of the entire drive system. Therefore, the motor drives require a proper solution to prevent current sensor faults and improve the reliability of the motor drive systems. The novelty of the proposed research includes integrating the current sensor fault-tolerant control (FTC) function according to enhanced technique into the field-oriented control loop for speed control of the motor drive system. Purpose. This research proposes a hybrid method involving a third difference operator and signal comparison algorithm to diagnose various types of current sensor faults as a positive solution to enhance the stability of the induction motor drive system. Methods. A hybrid method involving a third difference operator for the measured speed signals and a comparison algorithm between measured and estimated current signals are proposed to diagnose the current sensors’ health status in the fault-tolerant process. After determining the faulty sensor, the estimated current signals based on the Luenberger observer are used immediately to replace the defective sensor signal. Results. The current sensor is simulated with various failure types, from standard to rare failures, to evaluate the performance of the FTC method implemented in the MATLAB/Simulink environment. Simultaneously, a fault flag corresponding to a defective sensor should be presented as an indicator to execute the repair process for faulty sensors at the proper time. Practical value. Positive results have proven the feasibility and effectiveness of the proposed FTC integrated into the speed controller to improve reliability and ensure the stable operation of the induction motor drive system even under current sensor fault conditions. References 29, tables 3, figures 10.

A systematic review of fault tolerance techniques for smart city applications

Smart City Applications encompass many characteristics that increase the risk of failures, such as context-awareness, adaptiveness, distribution and heterogeneity. Therefore, it is important to implement fault-tolerant mechanisms to produce more reliable applications. This study presents a systematic literature review of fault tolerance techniques that have been proposed for, or applied to Smart City Applications. It also characterizes faults, errors and failures that may occur in these systems. To the best of our knowledge, this is the first review that provides a broad picture of the research area and points out research limitations and directions. We selected 43 primary studies and performed initial classifications (e.g., based on type of research, type of contribution, application domains and subdomains, and type of system architecture). We further classified and discussed the selected studies based on types of fault tolerance techniques and types of faults and failures. System Reconfiguration, Diversity, and Retry are classical techniques that have been investigated in this domain. Many fault and failure types have also been addressed. While those well-known techniques have been explored for introducing fault tolerance capabilities into Smart City Applications, others have been overlooked. Moreover, evidence on the effectiveness and applicability of the proposed fault tolerance solutions is still very limited.Editor’s note: Open Science material was validated by the Journal of Systems and Software Open Science Board.

An innovative NSGA-II-based Byzantine Fault Tolerant solution for software defined network environments

Byzantine fault tolerance (BFT) of the control plane in Software Defined Networking (SDN) is achieved by mapping each switch to 3f+1 number of controllers, where f represents the number of faulty controllers that can be tolerated at a time. A BFT approach protects the data plane from any potential malicious activity at the control plane by detecting the inconsistency among the response messages from multiple controllers. To compute the optimal mapping of switches to the controller, the existing literature does not consider some important parameters. This paper proposes a novel approach, named NBFT-SDN, that extends an artificial intelligence algorithm (i.e. NSGA-II) to solve a new formulated multi-objective optimization problem associated with this mapping. NBFT-SDN considers the very important parameters link reliability and link load along with switch-to-controller minimum delay, switch-to-controllers maximum reliability, controller-to-controller minimum delay, minimum link load, minimum hop count, and controller load balancing when mapping the switches to the controllers in optimum manner. The performance of our proposed approach is evaluated in comparison to a state-of-art approach using real network traces with network topologies of diverse sizes. Our proposed approach NBFT-SDN show improved network performance in terms of reliability, delay, hop count, load balancing and link load.

Fault Tolerance Conceptual Strategy for a Quadcopter Drone with Rotor Failure

The ability of a quadcopter drone to maintain its attitude relies solely on its four rotors. If even one motor fails, the drone loses its ability to hold attitude and altitude. This paper explores a new fault tolerance solution to enhance attitude control for quadcopter drones following the complete loss of a single rotor. By following the fundamental principle of balancing forces and moments on a quadrotor drone, the paper demonstrates that it is feasible to land the drone safely by minimizing roll, pitch, and yaw when a rotor fails. The concept centers around thrust vectoring, which allows an opposite motor to tilt independently. The results indicate that tilting the opposite rotor by 45o provides better management of the drone’s roll, pitch, and yaw, enabling the incapacitated drone to land in a more controlled and manageable manner. The paper includes simulation results and a summary table of the novel idea’s performance enhancements.

Self-stabilizing Byzantine fault-tolerant repeated reliable broadcast

We study a well-known communication abstraction called Byzantine Reliable Broadcast (BRB). This abstraction is central in the design and implementation of fault-tolerant distributed systems, as many fault-tolerant distributed applications require communication with provable guarantees on message deliveries. Our study focuses on fault-tolerant implementations for message-passing systems that are prone to process-failures, such as crashes and malicious behavior.At PODC 1983, Bracha and Toueg, in short, BT, solved the BRB problem. BT has optimal resilience since it can deal with t<n/3 Byzantine processes, where n is the number of processes. The present work aims to design an even more robust solution than BT by expanding its fault-model with self-stabilization, a vigorous notion of fault-tolerance. In addition to tolerating Byzantine and communication failures, self-stabilizing systems can recover after the occurrence of arbitrary transient-faults. These transient faults allow the model to represent temporary deviations from the assumptions on which the system was originally designed to operate (provided that the algorithm code remains intact).We propose, to the best of our knowledge, the first self-stabilizing Byzantine fault-tolerant (BFT) solution for repeated BRB in signature-free message-passing systems (that follows BT's problem specifications). Our contribution includes a self-stabilizing variation on BT that solves a single-instance BRB for asynchronous systems. We also consider the problem of recycling instances of single-instance BRB. Our self-stabilizing BFT recycling for time-free systems facilitates the concurrent handling of a predefined number of BRB invocations and, in this way, can serve as the basis for self-stabilizing BFT consensus.

The principle of operation of a decentralized money transfer information system

The article is devoted to reviewing the concept of blockchain in the implementation of a decentralized money transfer system. The concept of fault tolerance of cryptocurrency technology is considered and some aspects of the fault tolerance solution are given. The analysis of information security of enterprises was carried out. The questions of anonymity of banking systems are touched upon. The dependence of the anonymity of the banking system on the quality of the software and on the security policy is shown. Recommendations are given for solving the problem of anonymity, in particular, generating a private key or embedding an algorithm for connecting transactions. Approaches to creating transactions are considered. Examples of algorithms that allow protecting user transactions are given. The description of the blockchain structure and the consensus algorithm has been carried out. The issues of generating a chain of blocks of an honest user and a chain of blocks of an attacker are considered. Formalization of the estimate of the probability of transaction break-even has been carried out. Binomial random walk was used for estimation. The end of the transaction process is just as justified as waiting for new blocks to be added to the chain of an honest user. A digital signature is used to identify whether a transaction belongs to a public key. The processes of signing a transaction and verifying a signature are considered. Separately, the approaches used in the design of cryptocurrencies are highlighted. The network stores the entire history of transactions, it is possible to calculate the balance of the total amount of incoming funds and spending money. These approaches are software solutions based on game theory and cryptography. The main attention is drawn to the business values of cryptocurrencies and ways to achieve them. As a result, the necessary functionality of the software client is formulated. The article provides a thorough analysis of a new financial and settlement instrument - electronic money. The task of developing a virtual payment system has recently received much attention.

Energy-Aware Next-Generation Mobile Routing Chains with Fog Computing for Emerging Applications

The Internet of Things (IoT) provides robust services to connected sensors in a distributed manner, and maintains real-time communication using wireless standards. The smart network has offered many autonomous smart systems to collect information from remote nodes, and share it by exploring the network layer. Researchers have recently offered a variety of ways to increase the effectiveness of emerging applications using trustworthy relaying systems. However, there are still many issues with route reformulation due to frequent disconnections of mobile devices and resource limitations. Furthermore, most of the existing methods for IoT systems are unable to utilize network resources, which lowers the performance of green networks. Thus, providing a foolproof solution for the autonomous system with energy efficiency is a challenging task. Therefore, this paper presents an algorithm for the mobile network using fog computing to reduce network disconnectivity. Furthermore, using security services, the proposed algorithm efficiently explores the characteristics of the device, and avoids malicious traffic to drain the additional energy consumption of the network. The main aspects of the proposed algorithm are as follows: (i) using the adjustable transmission power, the proposed algorithm offers a fault-tolerant solution to transmit the aggregated data over the unpredictable wireless system; (ii) with the support of fog nodes, the data load is reduced among devices with the offering of a secured authentication scheme. Using simulations, the proposed algorithm is tested, and its significance is demonstrated against other related studies.

DRAGON: Dynamic Recurrent Accelerator for Graph Online Convolution

Despite the extraordinary applicative potentiality that dynamic graph inference may entail, its practical-physical implementation has been a topic seldom explored in literature. Although graph inference through neural networks has received plenty of algorithmic innovation, its transfer to the physical world has not found similar development. This is understandable since the most preeminent Euclidean acceleration techniques from CNN have little implication in the non-Euclidean nature of relational graphs. Instead of coping with the challenges arising from forcing naturally sparse structures into more inflexible stochastic arrangements, in DRAGON, we embrace this characteristic in order to promote acceleration. Inspired by high-performance computing approaches like Parallel Multi-moth Flame Optimization for Link Prediction (PMFO-LP), we propose and implement a novel efficient architecture, capable of producing similar speed-up and performance than baseline but at a fraction of its hardware requirements and power consumption. We leverage the hidden parallelistic capacity of our previously developed static graph convolutional processor ACE-GCN and expanded it with RNN structures, allowing the deployment of a multi-processing network referenced around a common pool of proximity-based centroids. Experimental results demonstrate outstanding acceleration. In comparison with the fastest CPU-based software implementation available in the literature, DRAGON has achieved roughly 191× speed-up. Under the largest configuration and dataset, DRAGON was also able to overtake a more power-hungry PMFO-LP by almost 1.59× in speed, and at around 89.59% in power efficiency. More importantly than raw acceleration, we demonstrate the unique functional qualities of our approach as a flexible and fault-tolerant solution that makes it an interesting alternative for an anthology of applicative scenarios.

Reliability-Driven Memristive Crossbar Design in Neuromorphic Computing Systems

In recent years, memristive crossbar-based neuromorphic computing systems (NCS) have provided a promising solution to the acceleration of neural networks. However, stuck-at faults (SAFs) in the memristor devices significantly degrade the computing accuracy of NCS. Besides, the memristor suffers from the process variations, causing deviation of the actual programming resistance from its target resistance. In this paper, we propose a reliability-driven design framework for a memristive crossbar-based NCS in combination with general and chip-specific design optimizations. First, we design a general reliability-aware training scheme to enhance the robustness of NCS to SAFs and device variations; a dropconnect-inspired approach is developed to alleviate the impact of SAFs; a new weighted error function, including cross-entropy error (CEE), the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$l_{2}$ </tex-math></inline-formula> -norm of weights, and the sum of squares of first-order derivatives of CEE with respect to weights, is proposed to obtain a smooth error curve, where the effects of variations are suppressed. Second, given the neural network model generated by the reliability-aware training scheme, we exploit chip-specific mapping and re- training to further improve computation accuracy loss incurred by SAFs. Experimental results show that the proposed method can boost the computation accuracy of NCS and improve the NCS robustness. Note to Practitioners—This work is motivated by the manufacturing reliability problem in a memristive crossbar-based NCS. To enhance the robustness of an NCS to SAFs and device variations, this paper presents a reliability-driven design framework with taking account of both general and chip-specific design optimizations. The experimental results have demonstrated that the proposed framework is superior to the prior arts, and can be easily integrated with existing industrial hardware-based fault tolerance solutions for higher accuracy at lower overhead. Memristive crossbar-based computing system gives hope for the anticipated efficient implementation of artificial neuromorphic networks. With the help of the reliability-driven designs, the computation accuracy is restored, and hence we can expect the wide use of memristive crossbar-based computing system in neuromorphic computing applications.

Recent Advances and Future Prospects of Using AI Solutions for Security, Fault Tolerance, and QoS Challenges in WSNs

The increasing relevance and significant acceptance of Wireless Sensor Network (WSN) solutions have aided the creation of smart environments in a multitude of sectors, including the Internet of Things, and offer ubiquitous practical applications. We examine current research trends in WSN using Artificial Intelligence (AI) technologies and the potential application of these methods for WSN improvement in this study. We emphasize the security, fault detection and tolerance, and quality of service (QoS) concerns in WSN, and provide a detailed review of current research that used different AI technologies to satisfy particular WSN objectives from 2010 to 2022. Specifically, this study’s purpose is to give a current review that compares various AI methodologies in order to provide insights for tackling existing WSN difficulties. Furthermore, there has been minimal existing related work concentrating employing AI approaches to solve security, fault detection and tolerance, and quality of service (QoS) concerns associated to WSN, and our goal is to fill the gap in existing studies. The application of AI solutions for WSN is the goal of this work, and we explore all parts of it in order to meet different WSN challenges such as security, fault detection and tolerance, and QoS. This will lead to an increased understanding of current AI applications in the areas of security, fault detection and tolerance, and QoS. Secondly, we present a comprehensive study and analysis of various AI schemes utilized in WSNs, which will aid the researchers in recognizing the most widely used techniques and the merits of employing various AI solutions to tackle WSN-related challenges. Finally, a list of open research issues has been provided, together with considerable bibliographic information, which provides useful recent research trends on the topics and encourages new research directions and possibilities.

Single-Phase Fault Tolerant Multilevel Inverter Topologies—Comprehensive Review and Novel Comparative Factors

Multilevel inverters (MLIs) are used in a variety of industrial applications in high- and medium-voltage systems. The modularity, high-power output from medium voltages, and low harmonic content are some of the advantages of MLIs. The reliability of MLIs is quite important. The reliability is affected by different kinds of faults occurring in the MLIs. In MLI circuits, switching devices are the most vulnerable components and have a major involvement in all types of faults. As an outcome, it is necessary to take proper corrective action in the event of a fault. This work provides a comprehensive review of different fault tolerant (FT) solutions for MLIs in the event of switch fault. Moreover, various single-phase FT MLI topologies are reviewed, along with their constructional features, merits, and demerits. This work also proposes a comparison approach that integrates novel factors to account for fault tolerance quantitatively. A comparison investigation verifies the effectiveness of the proposed method. The FT operation of an existing five-level FT MLI topology is discussed, simulated, and experimentally verified.

Three-Phase Current Reconstruction for PMSM Drive With Modified Twelve Sector Space Vector Pulse Width Modulation

To reduce the cost and size of the power converters, techniques of reconstructing three-phase current through a single current sensor have been employed in permanent-magnet synchronous motors. In existing studies, the reconstruction precision is largely affected by current reconstruction dead zones in the space vector pulsewidth modulation (PWM) plane, which requires additional algorithms to compensateeither by modifying PWM modulation strategy or by phase-shifting of PWM signal. In this article, a phase current reconstruction method for three-phase inverter with the modification of a twelve sector space vector pulsewidth modulation (TS-SVPWM) is proposed. By arranging a single current sensor on a circuit branch between two power switches, the proposed method can reconstruct the three-phase current of the voltage source inverter and avoid the dead zones located at the sector boundary region. This method can be realized not only with a single Hall-effect current sensor but also with a single shunt resistor, which can expand its application. The proposed method can be a fault-tolerance solution when there are phase current sensor faults. Beside, the three-phase current can be reconstructed when there are open-circuit faults in the power switches. Compared with previous methods, the proposed TS-SVPWM can reduce the phase current harmonics. The effectiveness of the proposed method is validated by experiments.

Verification of colorable hypergraph states with stabilizer test

Many-body quantum states, as a matter of fact, are extremely essential to solve certain mathematical problems or simulate quantum systems in measurement-based quantum computation. However, how to verify large-scale quantum states, such as hypergraph states, is an exceedingly hard task for many-body quantum systems. Here, we propose a novel fault-tolerant solution for the verification of colorable hypergraph states by using the stabilizer test. Furthermore, our protocol is dramatically facilitated by making only Pauli-X and Pauli-Z measurements. For geometric structure hypergraph states, the computational complexity of our protocol is polynomial. As to appliance, it will be also applied to blind quantum computing based on the no-signaling principle.

Fault-Tolerant Reconfiguration Topology and Control Strategy for Symmetric Open-Winding Multiphase Machines

This paper introduces a fault-tolerant reconfiguration topology and control strategy for symmetric open-winding multiphase machines. Bidirectional switching devices are proposed to connect the inverter bridge legs on different ends when the open-circuit fault occurs, and the reconfigured circuit can regenerate the missing voltage phasors due to the faults. Furthermore, to prevent overcurrent problems on the reconfigured bridge legs and maximize the torque output capability of the reconstructed system, winding currents are optimized with zero-sequence component injection during fault-tolerant operation. The experiments implemented on the five-phase system validate that the proposed fault-tolerant scheme is effective and has higher torque output capability compared with the traditional fault-tolerant solutions.

Unified Strategy for Fault-Tolerant Operation of MMC with Multiple SMs Failure Based on SMs Grouping Management

Modular multilevel converter (MMC) is distinguished by its modularity. In order to improve its reliability and avoid unscheduled maintenance, it requires the MMC to continue operating even though some of its submodules (SMs) have failed. In this paper, a grouping management method of SMs in MMC is first presented where every six SMs are conceptually grouped into a virtual subunit (SU), and on this basis, a simplified space vector modulation (SVM) implementation is developed. Then, the fault-tolerant solutions are proposed by investigating the post-fault operation of the MMC based on the virtual SU concept. It is analyzed that the SU failures can be categorized into three basic types, according to the phase where the failed SMs are located, and also multiple SM failures can be decomposed into one or multiple basic fault types. In this way, the fault-tolerant solutions are unified regardless of the number of faulty SMs. Further, the sorting capacitor voltage control is combined into the proposed methods to balance all of the SM capacitor voltages. With the proposed fault-tolerant method, the amplitude and THD of the line voltages are almost unchanged under fault conditions when compared with normal operations. The simulation and experiment verify the propositions.

Efficient Identification of Critical Faults in Memristor-Based Inferencing Accelerators

Deep neural networks (DNNs) are becoming ubiquitous, but hardware-level reliability is a concern when DNN models are mapped to emerging neuromorphic technologies such as memristor-based crossbars. As DNN architectures are inherently fault tolerant and many faults do not affect inferencing accuracy, careful analysis must be carried out to identify faults that are critical for a given application. We present a misclassification-driven training (MDT) algorithm to efficiently identify critical faults (FCFs) in the crossbar. Our results for three DNNs on the CIFAR-10 data set show that MDT can rapidly and accurately identify a large number of FCFs—up to <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$20\times $ </tex-math></inline-formula> faster than a baseline method of forward inferencing with randomly injected faults. We use the set of FCFs obtained using MDT and the set of benign faults obtained using forward inferencing to train a machine learning (ML) model to efficiently classify all the crossbar faults in terms of their criticality. Using the ground truth generated using MDT and forward inferencing, we show that the ML models can classify millions of faults within minutes with a remarkably high classification accuracy of up to 99%. We also show that the ML model trained using CIFAR-10 provides high accuracy when it is used to carry out fault classification for the ImageNet data set. We present a fault-tolerance solution that exploits this high degree of criticality-classification accuracy, leading to a 92.5% reduction in the redundancy needed for fault tolerance.

Fault-Tolerant Three-Phase VSI Based on a Modified Impedance Source Boost Inverter

This article proposes a fault-tolerant solution for the three-phase dc–ac converter using a modified impedance source boost inverter. Besides the additional modified impedance source, the solution is based on the reconfiguration of the topology to maintain the desired operation in case of an open-circuit failure mode of a power device. The proposed topology presents important features such as a boost capability to operate under fault condition and avoid extra active power devices in normal operation (without failures). Besides that, the circuit does not require dead times between the switches of each leg, which increases the reliability of the topology. Regarding the control solution for this converter, it is adopted a closed-loop current controller combined with a vectorial voltage modulator. Since this modulator is also function of the shoot-through, it will be possible to recover the amplitude voltage space vectors due to failures in power devices. The combination of these elements allows to obtain three-phase balanced ac currents with reduced total harmonic distortion even in fault-tolerant operation mode. The fault-tolerant performance of this solution is confirmed through several simulations and experimental results.

Wind Turbine Pitch Actuator Regulation for Efficient and Reliable Energy Conversion: A Fault-Tolerant Constrained Control Solution

Motivated for improving the efficiency and reliability of wind turbine energy conversion, this paper presents an advanced control design that enhances the power regulation efficiency and reliability. The constrained behavior of the wind turbine is taken into account, by using the barrier Lyapunov function in the analysis of the Lyapunov direct method. This, consequently, guarantees that the generated power remains within the desired bounds to satisfy the grid power demand. Moreover, a Nussbaum-type function is utilized in the control scheme, to cope with the unpredictable wind speed. This eliminates the need for accurate wind speed measurement or estimation. Furthermore, via properly designed adaptive laws, a robust actuator fault-tolerant capability is integrated into the scheme, handling the model uncertainty. Numerical simulations are performed on a high-fidelity wind turbine benchmark model, under different fault scenarios, to verify the effectiveness of the developed design. Furthermore, a Monte-Carlo analysis is exploited for the evaluation of the reliability and robustness characteristics against the model-reality mismatch, measurement errors and disturbance effects.