Autonomous Driving Systems (ADS) are critical components of Intelligent Transportation Systems (ITS), where vehicle-level reliability has a direct bearing on road traffic safety. Evaluating ADS performance in complex environments remains challenging due to the absence of test oracles and the heavy reliance on deep learning. To address these challenges, this study proposes a novel metamorphic testing framework tailored for vision-based ADS. First, causal inference is employed to extract key environmental factors from high-dimensional observational traffic data, thereby reducing the test space. Second, a multi-objective optimization algorithm integrating causal counterfactual reasoning is developed to quantify the challenges associated with specific combinations of causal factors, enabling cost-effective exploration of test conditions. Third, low-risk source images are systematically transformed into hazardous driving scenes through a fine-tuned diffusion model, allowing ADS evaluation to be guided by metamorphic relations (MRs). Empirical experiments show that the proposed method achieves a higher fault detection ratio than the strongest baseline in four out of five ADS models, with relative gains ranging from 18.1% to 88.9%. Data augmentation experiments further demonstrate that incorporating MR-violating test cases can reduce ADS prediction errors by up to 13.67%, with these benefits preserved in real-world road traffic datasets through domain adaptation. This study highlights a new pathway for validating the reliability of vision-based ADS driven by deep learning, thereby supporting the deployment of safer road transportation. The source code for our methods and baselines is available athttps://github.com/SafeDL/AutoMetTest

Large multimodal language models (MLLMs) have revolutionized natural language processing and visual understanding, but often contain outdated or inaccurate information. Current multimodal knowledge editing evaluations are limited in scope and potentially biased, focusing on narrow tasks and failing to assess the impact on in-domain samples. To address these issues, we introduce ComprehendEdit, a comprehensive benchmark comprising eight diverse tasks from multiple datasets. We propose two novel metrics: Knowledge Generalization Index (KGI) and Knowledge Preservation Index (KPI), which evaluate editing effects on in-domain samples without relying on AI-synthetic samples. Based on insights from our framework, we establish Hierarchical In-Context Editing (HICE), a baseline method employing a two-stage approach that balances performance across all metrics. This study provides a more comprehensive evaluation framework for multimodal knowledge editing, reveals unique challenges in this field, and offers a baseline method demonstrating improved performance. Our work opens new perspectives for future research and provides a foundation for developing more robust and effective editing techniques for MLLMs.

Gaussian process (GP) offers a robust solution for modeling the dynamics of unmanned surface vehicles (USV) in model-based reinforcement learning (MBRL). However, the rapidly increasing computational complexity with a large sample capacity of GP limits its application in complex scenarios that require substantial samples to cover the state space. In this article, a novel probabilistic MBRL approach, probabilistic neural networks model predictive control (PNMPC) is proposed to tackle this issue. With an iterative learning framework, PNMPC properly models the USV dynamics using neural networks from a probabilistic perspective to avoid the computational complexity associated with sample capacity. Employing this model to effectively propagate system uncertainties, a model predictive control (MPC) policy is developed to robustly control the USV against external disturbances. Evaluated by position-keeping and multiple targets-tracking scenarios on a real USV data-driven simulation, the proposed method consistently demonstrates its significant superiority in both model accuracy and control performance compared to not only GP model-based approaches but also the probabilistic neural networks-based MBRL baselines, across various scales of external disturbances.Note to Practitioners—Modelling the system dynamics and maintaining computational efficiency with a large sample set has been challenging for MBRL in the USV domain. We propose a novel neural network modeling method to capture the dynamic features of USV within an RL loop and develop a robust MPC policy based on its uncertainty propagation. Our method achieves computational complexity independent of the sample capacity and outperforms related baselines in model accuracy and control performance.

This paper focused on developing an effective Multi-Agent Reinforcement Learning (MARL) approach that quickly explores optimal control policies of multiple agents through interactions with unknown environments. Multi-Agent Continuous Dynamic Policy Gradient (MACDPP) was proposed to tackle the issues of limited capability and sample efficiency in the current MARL approaches. It alleviates the inconsistency of multiple agents’ policy updates by introducing the relative entropy regularization to the Centralized Training with Decentralized Execution (CTDE) framework with the Actor-Critic (AC) structure. Evaluated by multi-agent cooperation and competition tasks and traditional control tasks including OpenAI benchmarks and robot arm manipulation, MACDPP demonstrates its significant superiority in learning capability and sample efficiency compared with both related multi-agent and widely implemented signal-agent baselines. It converges to$62\%$higher average return and uses$38\%$fewer samples compared with the suboptimal baseline over all tasks, indicating the potential of MARL in challenging control scenarios, especially when the number of interactions is limited. The open source code of MACDPP is available at https://github.com/AdrienLin1/MACDPP.Note to Practitioners—Learning proper cooperation strategy over multiple agents in complicated systems has been a challenge in the domain of Reinforcement Learning. Our work extends the traditional MARL approach FKDPP that has been successfully implemented in the real-world chemical plant by Yokogawa to the CTDE framework and AC structure that supports continuous actions. This extension significantly expands its range of applications from cooperative/competitive tasks to the joint control of one complex system while maintaining its effectiveness.

This article addresses the robustness of unmanned surface vehicles (USV) using model-based reinforcement learning (MBRL). A novel MBRL approach, Lyapunov probabilistic model predictive control (LPMPC) is proposed to simultaneously learn both the probabilistic model of a USV and its corresponding estimated Lyapunov region of attraction (ROA) under one reinforcement learning framework. Unlike the existing MBRL USV systems with less consideration of robustness and safety, our method naturally learns a general indicator of system stability based on the probabilistic model’s belief and employs it to guide its policy. Evaluated by different navigation tasks in a simulation driven by real boat data, LPMPC demonstrated significant advantages in both control robustness and task completion against various levels of environmental disturbances compared with the baseline approach without Lyapunov ROA’s guidance. Note to Practitioners—Modelling the system stability without human prior knowledge is challenging in the domain of USV. This work proposed a data-driven method to iteratively learn a task-relevant stability model of USV in a probabilistic view. Based on the evaluation of a real boat data-driven simulation, the learned stability model contributed to superior driving skills in different USV scenarios by properly indicating and avoiding potentially risky states. In future research, we plan to expand the definition of risks in different tasks, such as loss of control, overlarge sway, and excessive energy consumption and investigate the proposed approach in real-world USV.

Automotive-grade chips play a crucial role in the development of intelligent networked vehicles, and functional safety is a key issue of automotive-grade chips. Transient faults induced by high-energy particle radiation affect the functional safety of automotive-grade chips significantly and may lead to catastrophic accidents. Currently, research on transient faults and chip functional safety, both domestically and internationally, remains relatively fragmented, with a clear lack of comprehensive functional safety analysis and testing methods for transient faults. Requirements for the analysis of transient faults are included in the ISO 26262 standard on applications to semiconductors. This paper integrates fault probability, circuit structure, and safety objectives into the analysis and testing of transient faults, enabling the accurate identification of vulnerable components solely related to functional safety during the design phase and quantifying their safety impact. During the testing phase, the thesis focuses solely on testing the vulnerable components identified during the design phase and prioritizes test patterns based on the contribution of each test pattern, allowing for the rapid identification of safety vulnerabilities through a limited number of test patterns in the testing phase. The experimental results demonstrate that our proposed method is capable of accurately detecting transient faults associated with safety. By utilizing only 452 screened and optimized test patterns, the test results have attained a coverage rate of 97.16%. This signifies a notable 56.3% enhancement in test efficiency, with a minimal and insignificant loss of just 1.7% in coverage.

Continual learning, focusing on sequential knowledge acquisition and retention, necessitates efficient memory management. This paper introduces a holistic approach, diverging from traditional methods that separately optimize neural network and replay buffer memory. We aim to enhance overall memory efficiency, addressing neural network parameters and replay buffer concurrently within strict memory constraints. This is achieved by harnessing neural network parameter redundancies and employing compression techniques like pruning and quantization, allowing data replay storage without extra memory overhead. Balancing memory use across components is challenging due to the complex search space of combined tasks. We tackle this by conceptualizing it as a bi-level optimization problem, integrating all tasks under a single objective, thus optimizing memory use and managing the interplay between different components. We employ a synergy of optimization techniques to solve this challenging bi-level optimization problem. Our experimental findings affirm the superior performance of our proposed method, outperforming existing techniques such as prompt-based, feature-replay, exemplar-replay, and regularization-based methods under stringent memory constraints, consistently across various datasets and neural network architectures.

In embedded systems, particularly resource-constrained Internet of Things (IoT) devices, the SM2 Digital Signature Algorithm (SM2-DSA) standard is widely deployed for cryptographic security. While fault injection attacks can compromise digital signatures and extract private keys without physical damage, traditional approaches require precise temporal or spatial control, resulting in limited success rates and revealing insufficient research into the potential vulnerabilities of SM2-DSA. To address this issue, this article introduces a novel and efficient lattice-based fault attack method targeting SM2-DSA. The method involves injecting faults into the nonce before the fourth step of the signature operation. By leveraging both the correct and erroneous intermediate values of Q obtained from the signature and verification processes, we can deduce partial bits of the nonce. Following this, we construct a lattice attack to recover the private key. Additionally, we establish the theoretical security boundary for lattice attack against SM2-DSA. Building upon the boundary, we propose an efficient implementation scheme for the attack. Experimental results demonstrate a 100% success rate over 1,000 trials, using 61 signatures with six known bits of nonces for 256-bit SM2-DSA, with each recovery process completed in under three seconds. Finally, we propose countermeasures against this attack. Our proposed attack reveals potential security vulnerabilities in SM2-DSA implementations, providing constructive guidance for enhancing algorithmic security measures and defensive countermeasures.

This article addresses the prediction stability, prediction accuracy, and control capability of the current probabilistic model-based reinforcement learning (MBRL) built on neural networks. A novel approach to dropout-based probabilistic ensembles with trajectory sampling (DPETS) is proposed, where the system uncertainty is stably predicted by combining the Monte Carlo dropout (MC Dropout) and trajectory sampling in one framework. Its loss function is designed to correct the fitting error of neural networks for more accurate prediction of probabilistic models. The state propagation in its policy is extended to filter the aleatoric uncertainty for superior control capability. Evaluated by several Mujoco benchmark control tasks under additional disturbances and one practical robot arm manipulation task, DPETS outperforms related MBRL approaches in both average return and convergence velocity while achieving superior performance than well-known model-free baselines with significant sample efficiency. The open-source code of DPETS is available at https://github.com/mrjun123/DPETS.

In this article, a novel reinforcement learning (RL) approach, continuous dynamic policy programming (CDPP), is proposed to tackle the issues of both learning stability and sample efficiency in the current RL methods with continuous actions. The proposed method naturally extends the relative entropy regularization from the value function-based framework to the actor-critic (AC) framework of deep deterministic policy gradient (DDPG) to stabilize the learning process in continuous action space. It tackles the intractable softmax operation over continuous actions in the critic by Monte Carlo estimation and explores the practical advantages of the Mellowmax operator. A Boltzmann sampling policy is proposed to guide the exploration of actor following the relative entropy regularized critic for superior learning capability, exploration efficiency, and robustness. Evaluated by several benchmark and real-robot-based simulation tasks, the proposed method illustrates the positive impact of the relative entropy regularization including efficient exploration behavior and stable policy update in RL with continuous action space and successfully outperforms the related baseline approaches in both sample efficiency and learning stability.

Test the safety of Autonomous Driving Systems (ADS) with realistic traffic conditions is important to the insurance industry, legislators, and third-party technical services. Approaches for ADS testing can be divided into two main categories: physical test and virtual test, as shown in Fig. 1.

Visual-based object detection has become a crucial component in the realm of autonomous vehicles. However, conducting reliable testing for such systems remains unresolved. In this paper, we advocate for the application of causal inference to investigate the pivotal environmental factors influencing detection accuracy. Through the integration of diffusion models, we address the specialized conditional generalization of hazardous testing images. Our approach involves the construction of observational data to attribute key factors and fine-tune the diffusion model. Additionally, we introduce an optimal prompt words search method that strikes a balance between test coverage and level of challenge. Subsequently, leveraging these optimal prompts, we propose a cost-effective testing image generation through both "Text2Scene" and "Image2Scene" fashions. The experimental results indicate that, on the generalized dataset, the performance of object detection algorithms is the poorest, with the average detection accuracy decreasing from 0.81 to 0.285. Moreover, retraining object detection models on our generalized critical test cases can ultimately enhance algorithm performance, achieving a median accuracy improvement of up to 8.13%. Overall, our research proposes a novel approach to generalize test cases, thereby contributing to the advancement and deployment of safer autonomous vehicles.

Reinforcement learning has been shown to have great potential applications in autonomous driving. For collaborative driving scenarios, multi-agent reinforcement learning can be used to explore efficient and collaborative driving strategies. However, it still faces the challenge of non-stationary. Traditional methods focus on evaluating the similarities between the real state of the teammate and the modeled state. There is also the issue of partial observability. It can be addressed by establishing communication to share information with other surrounding agents. However, prior approaches overlook the efficient communication problem caused by unprocessed and redundant information. To tackle these two challenges, we propose an approach named Multi-Agent Collaboration via Cross-Attention and Communication (MACAC). MACAC leverages the agent’s local observations to analyze and capture environment and interaction information, while also incorporating teammate modeling through the exchange of concise state information via communication. In addition, to improve the learning process, we integrate the noisy advantage technique into MACAC to enhance the agent’s exploration capabilities. As a result, vehicles can effectively adapt to dynamic environments and exhibit efficient collaborative driving skills. In all, experiments conducted on an autonomous driving simulator demonstrate that our approach surpasses the performance of the baseline algorithms.

In this article, we focus on the computational efficiency of probabilistic model-based reinforcement learning (MBRL) in unmanned surface vehicles (USV) under unforeseeable and unobservable external disturbances. A novel MBRL approach, local update spectrum probabilistic model predictive control (LUSPMPC), is proposed to fully release the superiority of the probabilistic model approximated in the frequency domain in computational efficiency while mitigating its risk of overfitting during the learning procedure. It employs a local update strategy to relieve the violation of Bochner's theory, and a frequency clipping trick to encourage the approximated model to focus on the features in the low-frequency domain. Evaluated by the position-keeping task in a real USV data-driven simulation, LUSPMPC shows its significant advantages in computational efficiency while achieving better learning capability, generalization capability, and control performances in a wide range of sparse scales compared with the baseline MBRL approaches that approximate their models in sample space and frequency domain, and therefore becomes an appealing solution for MBRL USV system defending against rapidly changing ocean disturbances.

Automated driving systems (ADS) have made remarkable progress in recent years, yet their reliability and testability remain as significant challenges. The environmental conditions that ADS face are highly complex and may result in the disruption of autonomous vehicles. In this study, we propose an approach that leverages causal inference theory to analyze the impact of causal factors on automated driving visual modules. Our method uncovers the root key factors that affect visual perception performance. We further establish a Challenging Index to quantitatively characterize the causal effects of the key factors on perception failures. This quantitative index is subsequently utilized to generate risky scenarios. Through extensive experiments on various state-of-the-art automated driving visual algorithms, we demonstrate the effectiveness of the challenge index in evaluating the level of hazard in the deployment environment. Additionally, the proposed “challenge index guided search” method improves test efficiency by up to 8.95 times compared to the baselines while maintaining a balance between coverage diversity and the hazardous level of test scenarios. Our research offers a new perspective for analyzing and evaluating the impact of key factors on visual perception. This contributes to the reduction of test space and efficiency of the generation of high-value test scenarios, ultimately advancing the deployment of safer automated vehicles.

The rapid development of the Internet of Vehicles (IoV) provides a strong technical guarantee for intelligent transportation, greatly facilitating people's daily travel. At the same time, its security problems are becoming increasingly prominent. Fortunately, the cryptographic integrated circuits (ICs) provide a security guarantee for the IoV access authentication, traffic management, data communication, etc., which is the core and cornerstone of the IoV cryptographic technology. However, the IoV has high requirements for timeliness, and its communication resources are precious, so its necessary to ensure that the overhead of the cryptographic module is small and the delay is low. In this paper, we adopt SM2 Elliptic Curve Public Key Digital Signature Algorithm-with fast operation speed and short signature data to implement cryptographic ICs. We design and optimize its hardware design to balance overhead and efficiency. Based on the Montgomery point multiplication algorithm in Lopez-Dahab (LD) projection coordinates, we have researched the core point multiplication operation in SM2 and optimized the time-consuming operation in finite fields, which improved the computational efficiency of SM2. Finally, we completed the hardware design of SM2 on GF(2233) domain and verified it on a Xilinx Kintex-7 FPGA development board. The experiment results show that the design occupies a total of 77,665 Slice LUTs and merely takes 2.57 μs to complete a signature verification. The signature verification rate is 389,105 times/s. Compared with traditional solutions, our proposed method achieves less overhead and little latency.

In this article, we address the difficulty of controlling unmanned surface vehicles (USVs) under unforeseeable and unobservable external disturbances using model-based reinforcement learning (MBRL) without human’s prior knowledge. A novel MBRL approach, filtered probabilistic model predictive control (FPMPC) is proposed to iteratively learn the USV model and an MPC-based policy in a probabilistic way through trial-and-error interactions. Compared with existing MBRL approaches that model the unobservable disturbances as system noise, FPMPC introduces a Bayesian filter process to implicitly translate the system dynamics to a partially-observed Markov decision process to present those disturbances as hidden states. An adaptive sample selection is proposed to remove the redundant learning samples based on the filter belief. Equipped with bias compensation and parallel computation, an FPMPC system, specific for USV, is developed. Evaluated by both position holding and target reaching tasks in a real USV data-driven simulation, FPMPC shows its significant superiority in control performances, generalization capability, and sample efficiency under large disturbances compared with the baseline approaches.

Understanding and modeling human driver behavior and subsequently applying these patterns in various scenarios is crucial for autonomous vehicle motion planning. However, solution that naturally encodes human driving skills remains challenging due to the difficulty of balancing the variability of human behavior and the robustness in various driving environments. To tackle this issue, a novel motion planning approach is proposed based on Kernelized Movement Primitive (KMP) in this paper to adaptively learn human driving behavior in a stochastic way by employing Gaussian mixture model(GMM) and Gaussian mixture regression(GMR). The Kullback-Leibler(KL) divergence is utilized to minimize the information loss between imitating reference behavior and adapting new tasks and therefore generates robust motion trajectories in a variety of driving situations. The proposed approach is evaluated by a mature urban driving simulator CARLA. The experimental results shows its capability of generating robust driving trajectories by naturally adapting human driving skill into various driving situations.

Current incremental learning is confronted with the problem of catastrophic forgetting. Existing work usually addressed this problem by enlarging the sample database. But few people pay attention to the order-sensitive problem in incremental learning. In this article, we propose a task sorting method with the feature similarity between the consecutive tasks. Experimental results on CIFAR-100 and CORe50 datasets demonstrate that the learning sorting matters for incremental learning. The task sorted with the highest feature similarity will get a better performance than that of random sorting.

In this paper, a model predictive control (MPC) approach that combines sparse spectrum Gaussian processes model and unscented Kalman Filter is proposed for path tracking task in autonomous driving. To tackle the difficulty of balancing control performance and computational cost in MPC with Gaussian processes model, the proposed approach employs the sparse spectrum Gaussian processes (SSGP) to efficiently model the vehicle, and utilizes unscented Kalman filter (UKF) to naturally propagate model uncertainties during multiple step prediction of MPC. The proposed approach is evaluated in both a numerical driving simulation and a mature driving simulation CARLA. The results indicate that the proposed method achieves a robust driving performance with a significant reduction of computational complexity.

In order to effectively solve the problem of loop closures detection matching speed and loop closures false positives in large-scale maps, this paper proposes a loop closures detection method combining geomagnetic sequence search and lidar point cloud matching. By adding the geomagnetic matching algorithm to the loop closures, the candidate loop detection pose-node set is filtered, which reduces the false detection caused by the high local similarity in the lidar Simultaneous Localization and Mapping (SLAM), as well as the loop false positives and mapping distortion caused by the interference of reflection and transmission of laser beam. In this work, the performance of the algorithm is verified by the lidar point cloud and geomagnetic signal dataset collected in the real environment. The experimental results show that compared with the current product-level SLAM system, Cartographer, the proposed algorithm improves the loop detection speed by 31% (in data of 100 matches) and the matching accuracy is improved by 10% at the 20% recall.

In the outdoor dynamic scene, the current Monocular Visual odometry methods of simultaneous localization and mapping system (SLAM) have Low utilization of image information, and they are not enough to meet the problem of stable image matching tracking. An information fusion method combining feature points and direct method gray matching method is proposed, which not only has no serious feature loss when demanding in fast motion, but also reduces the dependence on gray-scale invariant hypothesis. The proposed method also has better stability in the scenes with insufficient texture information, and the method reduces the dependence on gray-scale invariant hypothesis. Experimental results show that our novel front-end visual odometry simplifies the feature points, enriches the front-end utilization information, improves the pixel information of the image, and improves the feature points loss and excessive dependence on the intensity. The average real-time frame rate is around 32Hz, meeting the requirements of real-time performance. In the large-scale scene of the KITTI visual odometry datasets, the jitter effect of the dynamic scene is weakened, and the root mean square error of the positioning is 1. 79m, The average root mean square error of the method is reduced to 22.77% of ORB-SLAM, 26.48% of Semi-Direct Monocular Visual Odometry(SVO),and 37.68% of Large-scale direct SLAM(LSD).

Planning a safe and comfortable trajectory in complex traffic scenarios is very challenging because there are many constraints to consider, such as vehicle dynamics constraints and traffic rules. The existing method is either to search for the trajectory in the lattice space, or to combine the front-end coarse path searching and the back-end trajectory smoothing. These methods only constrain the position, slope and second derivative of the trajectory externally, so that the planned trajectory is either too conservative to play the vehicle's motion performance or is so aggressive that the controller cannot track. We propose a motion planning method based on vehicle dynamics model prediction to solve an optimization problem involving vehicle dynamics constraints, safety and comfort requirements. Simulation results demonstrate that the proposed method can plan safe and smooth overtaking trajectory. Also it has good real-time performance and can run stably at 15 Hz.

Scale variation of a target frequently appears in tasks of visual tracking. Accurate scale estimation is challenging due to deformation, occlusion, rotation, change in the view angle and diversity of tracking object categories. Most tracking methods employ an exhaustive search of scales to estimate the target scales. However, only finite and discrete scales are usually searched due to the expensive computation requirement. Here, the authors propose a novel scale estimation method based on bounding box regression (BBR). They first formulate the scale tracking as a regression problem, and search for the entire continuous scale space without being limited by a manually specified number of scales. Then they extend the original single‐channel BBR to multi‐channel situations, to allow for better employment of multi‐channel features. To further take advantage of the time prior information of training samples, they derive a time‐related sample weighted multi‐channel BBR. Besides, they propose a quantitative measurement, scale divergence degree, to reflect the diversity of sampling strategy. Experimental results on OTB‐2015D dataset demonstrate that the proposed approach achieves outstanding scale estimation performance for visual tracking with significant deformation.

With the introduction of Connected and Autonomous Vehicles (CAVs), it is possible to extend the limited horizon of vehicles on the road by collective perceptions, where vehicles periodically share their sensory information with others using Vehide-2-Vehicle (V2V) communications. This technique relies on a certain number of participants to have a measurable advantage. Nevertheless, the spread of CAVs will take a considerable period of time, it is critical to understand the benefits and limits of V2V based collective perceptions in different market stages. In this work, we characterise the effective Field of View (eFoV) of a vehicle as the perception range using local sensors only, and the collective Field of View (cFoV) as the region learn from the network. Applying analytic and simulation studies in highway scenarios, we show that the eFoV drops quickly with the increase in traffic density due to blockage effects of surrounding vehicles, and it is insufficient to overcome this problem by increasing the sensing range of local sensors. On the other hand, vehicles can gain around 16 folds more information about the road environment by leveraging collective perceptions with only 10% CAV penetration rate. When the penetration rate reaches to around 30%, collective perceptions can provide 95% coverage over the road environments. Our analyses also show that apart from the benefits, employing collective perceptions could result in heavy broadcast redundancy, hence wasting the already scarce network resources. This observation suggests that the sharing of sensory information should be controlled appropriately to avoid overloading the communication networks.

Cryptographic ICs have been widely applied to numerous security-critical environments nowadays. Fault injection has become a serious attack on cryptographic IC, especially soft-errors or single event upsets (SEUs) by fine-resolution fault injection attacks. Detection and tamper evidence of these attacks become important. Traditional SEU diagnose methods usually require special sensors embedded into the circuits. However, these methods require non-trivial design and test effort, and usually just yield statistic results. In this paper, we formulate the detection fault injection attacks as a compressed sensing problem, due to sparsity of soft errors. Besides, due to the binary characteristic of the coefficient matrix and the variables, integer linear programming is adopted to reconstruct the soft error signals. Simulation results on a cryptographic IC demonstrate that the proposed method is capable to accurately detect the locations of soft-errors caused by fault injection attacks with negligible hardware overhead. The abnormal test output of scan-chains can be tamper evidence of the fault injection attacks.

The nanoscale CMOS technology has encountered severe reliability issues, especially in the on-chip memory. As the technology nodes keep shrinking, single-event upsets (SEUs) may encounter more frequent multiple-bit upsets (MBUs) per particle strike. The commonly used memory error correction methods, such as the single-error correction-double-error detection (SEC-DED) code, are no longer feasible. While the counterparts for multiple error correction codes (MECs) yield too costly overhead on delay and data redundancy, especially for MBUs in a word or a character, even with all bits upset, the error correcting ability of the existing error correcting methods is exceeded. In this paper, we introduce a novel block-based error detection and correction method for memory by analyzing the similarity of data. This method of error location and correction can cope with both single-word error (SWE) and multiple-word error (MWE), no matter how many corrupted bits of each word there are. Experimental results on the test system based on SRAM demonstrate that the proposed method can correct single-word-single-bit (SWSB), single-word-multiple-bit (SWMB), multiple-word-single-bit (MWSB), and multiple-word-multiple-bit (MWMB) errors. The proposed method based on block level greatly improved the correction ability with low redundancy, low decoding delay, and moderate complexity versus other homogeneous byte-level or bit-level protection methods. In particular, the detection and correction efficiency is more evident when the data size increases.

It is popular to develop the city-wide parking guidance system(PGS) in China nowadays, in order to alleviate the parking difficulties arising in large cities. Prediction on parking occupancy is the essential intelligent technology to help vehicles find the proper parking lot efficiently in PGS. And the known prediction methods have to be powered by real-time data, without which would cause significant inaccuracies. In the early stage of PGS deployment, however, it is very hard to collect the real-time data from the parking lots all over the city, considering the financial and time cost. So how could PGS try to keep the prediction on parking occupancy working well under the environment lacking realtime data? In this paper, we propose a method named SAW(non-stationary Stochastic And Wavelet neural network), to predict the parking occupancy at the given time, based on digging the history data. In the simulation, we compare our model with the largest Lyapunov exponents method and traditional wavelet neural network(WNN) by experimental tests using the same data, and the comparative analysis shows that the proposed model can effectively improve the long-term forecasting accuracy and achieve satisfactory results based on the smaller computational complexity.

Fast and efficient pedestrian detection and vehicle detection has become an increasingly important task in the autonomous driving technology. In this paper, we propose a new pedestrian detection and vehicle detection algorithm based on the YOLOv2 with optimized feature extraction. We adopt the priori experience about the feature box sizes, instead of K-mean clustering algorithm in the original YOLOv2 algorithm. We first conduct statistical analysis on the dataset with a label of pedestrian label and vehicles, and then we design the initial value of the pre-selection box that is more in line with the characteristics of pedestrian and vehicle. Together with hard negative mining, multi-scale training, and model pretraining, the proposed algorithm not only improves the detection accuracy but also keeps the good detection efficiency. Experimental results on traffic benchmark record demonstrate that the optimized algorithm satisfies the real-time capability and the accuracy requirement of the lowspeed autonomous driving.

Deep Neural Networks (DNNs) push the state-of-the-art in many machine learning applications, they often require millions of expensive floating-point operations for each input classification. This computation overhead limits the applicability of DNNs to low-power, embedded platforms and incurs high cost in data centers. In this paper, by exploiting the gap between the level of accuracy required by the applications/users and that provided by the computing system, we propose a model to balance the resource and quality of neural network with principal components analysis (PCA), which simplifies network structure (such as the number of input nodes and the number of network layers) by reducing the dimension of dataset and achieve diverse optimizations. The continuous iteration and optimization of the balance model can acquire the best balance of quality and resources of the neural network. As a result, the consumption of network resources is reduced to the greatest extent under the condition of satisfying the quality requirements. In this paper, a pedestrian recognition network is taken as an example. The results illustrate that the total number of network nodes is reduced by 69% while the recognition rate achieved (95.99%) can reach required level of recognition rate (95%), which is only 2% lower than the highest recognition rate (97.01%) of the original neural network of pedestrian recognition.

Single-event transients (SETs) have seriously deteriorated the reliability of integrated circuits (ICs), especially for those in mission- or security-critical applications. Detecting and locating SETs can be useful for fault analysis and design enhancement. Traditional methods of location identification of SETs usually require special sensors embedded into the circuits, or radiation scanning with fine resolutions over the surface for inspection. In this paper, we propose a method of location identification of SETs without sensors or image processing. We formulate location identification of SETs as a compressed sensing problem due to the sparsity and noncoherence observation. The simulation result on a cryptographic IC is demonstrated. The results illustrate that the proposed method has three advantages, compared to traditional SETs test methods: 1) the SETs sensitive area can be accurately identified; 2) the sampling rate is reduced by 64%; therefore, the test efficiency is largely enhanced with negligible hardware overhead; and 3) the method of location identification of SETs is robust to noise interference.

Fault injection attacks have constituted a serious threat against cryptographic integrated circuits (ICs). However, the security test nowadays is just sample test with workload statistics and experiences as the qualitative criterion, and results in costly, time‐consuming and error‐prone test procedures. This study presents a design for security test (DFST) method for cryptographic ICs against fault injection attacks. The DFST involves identifying the sensitive registers for various crypto modules, inserting the scan chains and generating the specific test patterns for security test. Then the security test is conducted on the manufactured cryptographic ICs with the industrial automatic test equipment. With this DFST method, a fast and automatic security test can be applied onto volume production of cryptographic ICs. Experimental results on an RSA implementation demonstrate the validity of this method.

BACKGROUND: The nonparametric trend test (NPT) is well suitable for identifying the genetic variants associated with quantitative traits when the trait values do not satisfy the normal distribution assumption. If the genetic model, defined according to the mode of inheritance, is known, the NPT derived under the given genetic model is optimal. However, in practice, the genetic model is often unknown beforehand. The NPT derived from an uncorrected model might result in loss of power. When the underlying genetic model is unknown, a robust test is preferred to maintain satisfactory power. RESULTS: We propose a two-phase procedure to handle the uncertainty of the genetic model for non-normal quantitative trait genetic association study. First, a model selection procedure is employed to help choose the genetic model. Then the optimal test derived under the selected model is constructed to test for possible association. To control the type I error rate, we derive the joint distribution of the test statistics developed in the two phases and obtain the proper size. CONCLUSIONS: The proposed method is more robust than existing methods through the simulation results and application to gene DNAH9 from the Genetic Analysis Workshop 16 for associated with Anti-cyclic citrullinated peptide antibody further demonstrate its performance.

TSVs can be fabricated with pitch of only tens of μm, and smaller. They can be densely distributed as inter-die interconnect in 3D ICs. However, the huge mismatch between the probe technology, such as the pitch of probe head and the capacity of probe card, and the TSV fabrication technology leads to an insufficient probe on TSV tips. In this paper, we present a novel TSV probing technique that can temporally bond pre-bond die to test interposer using anisotropic conductive adhesive material. On the two sides of the test interposer, TSVs and probe heads make contact with microbumps and C4-bumps, respectively. These two types of bumps are connected using redistribution metal layers, passing through the test interposer, which can bridge the gap between feature sizes of TSVs and probe head. This probing technology is also able to increase the test bandwidth by enlarging the test interposer and redistributing test signals between microbumps and C4-bumps. Moreover, the number of probe-card touchdown can be reduced by sharing the test interposer among multiple dies during the wafer-level testing. Simulation results on the corresponding test structures for TSVs open fault and leakage fault show the great test resolution and robustness considering different design choices and variable design parameters among the test structures.

Industrial process automation is undergoing an increased use of information communication technologies due to high flexibility interoperability and easy administration. But it also induces new security risks to existing and future systems. Intrusion detection is a key technology for security protection. However, traditional intrusion detection systems for the IT domain are not entirely suitable for industrial process automation. In this paper, multiple models are constructed by comprehensively analyzing the multidomain knowledge of field control layers in industrial process automation, with consideration of two aspects: physics and information. And then, a novel multimodel-based anomaly intrusion detection system with embedded intelligence and resilient coordination for the field control system in industrial process automation is designed. In the system, an anomaly detection based on multimodel is proposed, and the corresponding intelligent detection algorithms are designed. Furthermore, to overcome the disadvantages of anomaly detection, a classifier based on an intelligent hidden Markov model, is designed to differentiate the actual attacks from faults. Finally, based on a combination simulation platform using optimized performance network engineering tool, the detection accuracy and the real-time performance of the proposed intrusion detection system are analyzed in detail. Experimental results clearly demonstrate that the proposed system has good performance in terms of high precision and good real-time capability.

Testing through-silicon-vias (TSVs) is challenging largely due to the dimension gap between the TSVs and the probe needles. This paper proposes a low-cost and efficient test and diagnosis scheme without extra design for test structure or special probe technologies. A test probe head with current technology is in use, contacting multiple TSVs simultaneously. We propose an efficient binary search-based algorithm to guide the probe card movement and diagnose the faulty TSVs. To evaluate the test efficiency, mathematical analyses are performed considering geometric, probabilistic, and electronic issues including process variations and different probe architectures. The analysis demonstrates that the test efficiency can be largely enhanced by choosing appropriate sizes of probe needles, and by adjusting iteration algorithms according to the distribution of faulty TSVs.

Three-dimensional (3D) integrated circuits (ICs) that stack multiple dies vertically using through-silicon vias (TSVs) have gained wide interests of the semiconductor industry. The shift towards volume production of 3D-stacked ICs, however, requires their manufacturing yield to be commercially viable. Various techniques have been presented in the literature to address this important problem, including pre-bond testing techniques to tackle the “known good die” problem, TSV redundancy designs to provide defect-tolerance, and wafter/die matching solutions to improve the overall stack yield. In this paper, we survey recent advances in this filed and point out challenges to be resolved in the future.

ABSTRACT Laser fault injection has been proved to be a useful tool for attacks on integrated circuits. Transistors hit by a pulse of photons causes them to conduct transiently, thereby introducing transient logic errors, such as register value modifications, memory dumping, and so on. Attackers can make use of this abnormal behavior and extract sensitive information that the devices try to protect. This paper demonstrates laser fault injection attacks on very‐large‐scale integration circuits in a semi‐invasive way for the purpose of validating fault tolerant design and performance. Then, the paper presents a simulation methodology to evaluate the dependability of the integrated circuit design against laser fault injection attacks at design time. This simulation methodology involves exhaustively scanning the layout, incorporating the exposed cells into a circuit simulator, and examining the response of the circuit in detail. Experiments conducted on the same test chip spot the same vulnerabilities, thus indicating the validity of the proposed simulation methodology. Copyright © 2011 John Wiley & Sons, Ltd.

ABSTRACT To secure parallel systems in communication networks, in this paper, we propose a fast and scalable parallel scalar multiplication method over generic elliptic curves for elliptic curve cryptosystems, by means of our proposed scalar folding and unfolding techniques. In contrast to previous parallel scalar multiplication methods, our method can be implemented into scalable parallel computers. The optimal time complexity is k point doublings (D) plus log k point additions (A), denoted as kD + (log k)A, where k is the bit length of the scalar. If our method is applied to Koblitz curves, the optimal time complexity can be reduced to (log k)A. Furthermore, previous simple side‐channel‐protected scalar multiplication methods can be integrated into our method for resisting against simple side‐channel attacks. Copyright © 2011 John Wiley & Sons, Ltd.

Numerous research efforts have been made recently on exploiting the randomness of wireless channels to generate secret keys. This channel state information (CSI) based key generation method relies on spatial independence between the legitimate and eavesdropping channels. In this paper, we propose a methodology to extract secret keys through side-channel information. Our methodology involves the capture of side-channel information leaked by the electronic instruments, either when measuring channel characteristics or during encryption/decryption for key confirmation. Secret keys are extracted via analysis of the side-channel information. We provide experiment results to demonstrate the feasibility of our proposed side-channel attack methodology.

Abstract Power analysis attacks pose a serious threat to the security of many trusted systems. The principle of power analysis attacks is based on the assumption that the power consumption of an electronic device is proportional to the Hamming weight (HW) of the data being processed. However, this power model is defective as it is deviated from the CMOS circuit power consumption theory where power consumption is largely dependent on the switching ability, i.e., the Hamming distance (HD) of the data being processed. This paper presents an HD power analysis model which emulates Hamming distance based on probability distribution of HW, thus conjointly achieving better feasibility and accuracy. The experiment of CPA analysis on smart card chips running DES (Data Encryption Standard) and AES (Advanced Encryption Standard) encryption demonstrates that the proposed model can achieve 10% ∼ 18% better results compared to the existing HW model, which suggests higher success rate of discerning the secret key from the trusted systems. Copyright © 2010 John Wiley & Sons, Ltd.

The scalar multiplication is the dominant operation in Elliptic Curve Cryptosystems (ECC). It consists of a series of point additions and point doublings. A number of algorithms have been proposed to accelerate the scalar multiplication. Most of the algorithms demand high complexity which makes scalar multiplication hard to implement. In this paper, we propose an efficient algorithm for computing scalar multiplication based on partitioning scalar and propositional logic theory to address the trade-offs between speed and complexity. Our algorithm remains low complexity compared to existing accelerated scalar multiplication algorithms, whilst it is suitable for parallel processing systems.