Umeå University's logo

umu.sePublikasjoner
Endre søk
Begrens søket
1 - 37 of 37
RefereraExporteraLink til resultatlisten
Permanent link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Treff pr side
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sortering
  • Standard (Relevans)
  • Forfatter A-Ø
  • Forfatter Ø-A
  • Tittel A-Ø
  • Tittel Ø-A
  • Type publikasjon A-Ø
  • Type publikasjon Ø-A
  • Eldste først
  • Nyeste først
  • Skapad (Eldste først)
  • Skapad (Nyeste først)
  • Senast uppdaterad (Eldste først)
  • Senast uppdaterad (Nyeste først)
  • Disputationsdatum (tidligste først)
  • Disputationsdatum (siste først)
  • Standard (Relevans)
  • Forfatter A-Ø
  • Forfatter Ø-A
  • Tittel A-Ø
  • Tittel Ø-A
  • Type publikasjon A-Ø
  • Type publikasjon Ø-A
  • Eldste først
  • Nyeste først
  • Skapad (Eldste først)
  • Skapad (Nyeste først)
  • Senast uppdaterad (Eldste først)
  • Senast uppdaterad (Nyeste først)
  • Disputationsdatum (tidligste først)
  • Disputationsdatum (siste først)
Merk
Maxantalet träffar du kan exportera från sökgränssnittet är 250. Vid större uttag använd dig av utsökningar.
  • 1.
    Chang, Shuangshuang
    et al.
    School of Computer Science and Engineering, Northeastern University, Shenyang, China.
    Bi, Ran
    School Of Computer Science and Technology, Dalian University of Technology, Dalian, China.
    Sun, Jinghao
    School Of Computer Science and Technology, Dalian University of Technology, Dalian, China.
    Liu, Weichen
    School of Computer Science and Engineering, Nanyang Technological University, Singapore, Singapore.
    Yu, Qi
    School Of Computer Science and Technology, Dalian University of Technology, Dalian, China.
    Deng, Qingxu
    School of Computer Science and Engineering, Northeastern University, Shenyang, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Towards minimum WCRT bound for DAG tasks under prioritized list scheduling algorithms2022Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151, Vol. 41, nr 11, s. 3874-3885Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Many modern real-time parallel applications can be modeled as a directed acyclic graph (DAG) task. Recent studies show that the worst-case response time (WCRT) bound of a DAG task can be significantly reduced when the execution order of the vertices is determined by the priority assigned to each vertex of the DAG. How to obtain the optimal vertex priority assignment, and how far from the best-known WCRT bound of a DAG task to the minimum WCRT bound are still open problems. In this paper, we aim to construct the optimal vertex priority assignment and derive the minimum WCRT bound for the DAG task. We encode the priority assignment problem into an integer linear programming (ILP) formulation. To solve the ILP model efficiently, we do not involve all variables or constraints. Instead, we solve the ILP model iteratively, i.e., we initially solve the ILP model with only a few primary variables and constraints, and then at each iteration, we increment the ILP model with the variables and constraints which are more likely to derive the optimal priority assignment. Experimental work shows that our method is capable of solving the ILP model optimally without involving too many variables or constraints, e.g., for instances with 50 vertices, we find the optimal priority assignment by involving 12.67% variables on average and within several minutes on average.

  • 2. De Alcantara Dias, Bruno Martin
    et al.
    Maria Lagana, Armando Antonio
    Justo, Joao Francisco
    Yoshioka, Leopoldo Rideki
    Dias Santos, Max Mauro
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Model-based development of an engine control module for a spark ignition engine2018Inngår i: IEEE Access, E-ISSN 2169-3536, Vol. 6, s. 53638-53649Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    A Spark ignition (SI) engine is a complex, multi-domain component of the vehicle powertrain system. The engine control module (ECM) for an SI engine must achieve both high performance and good fuel efficiency. In this paper, we present a model-based development methodology for an open architecture ECM, addressing the entire development lifecycle including a control algorithm design, parameter calibration, hardware/software implementation, and verification/validation of the final system, both with bench tests on a dynamometer and in a real vehicle on the road. The ECM is able to achieve similar performance as the original proprietary ECM provided by the original equipment manufacturer. Its flexible and modular design enables easy extensibility with new control algorithms, and development of new engine types.

  • 3.
    Ding, Feng
    et al.
    School of Management, Nanchang University, Nanchang, Jiangxi 330031, China..
    Yu, Keping
    Global Information and Telecommunication Institute, Waseda University, Tokyo 169-8555, Japan (e-mail: keping.yu@aoni.waseda.jp).
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Li, Xiangjun
    School of Software, Nanchang University, Nanchang 330047, China..
    Shi, Yunqing
    Department of Electrical and Computer Engineering, New Jersey Institute of Technology, Newark, NJ 07101 USA..
    Perceptual Enhancement for Autonomous Vehicles: Restoring Visually Degraded Images for Context Prediction via Adversarial Training2022Inngår i: IEEE transactions on intelligent transportation systems (Print), ISSN 1524-9050, E-ISSN 1558-0016, Vol. 23, nr 7, s. 9430-9441Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Realizing autonomous vehicles is one of the ultimate dreams for humans. However, perceptual information collected by sensors in dynamic and complicated environments, in particular, vision information, may exhibit various types of degradation. This may lead to mispredictions of context followed by more severe consequences. Thus, it is necessary to improve degraded images before employing them for context prediction. To this end, we propose a generative adversarial network to restore images from common types of degradation. The proposed model features a novel architecture with an inverse and a reverse module to address additional attributes between image styles. With the supplementary information, the decoding for restoration can be more precise. In addition, we develop a loss function to stabilize the adversarial training with better training efficiency for the proposed model. Compared with several state-of-the-art methods, the proposed method can achieve better restoration performance with high efficiency. It is highly reliable for assisting in context prediction in autonomous vehicles.

  • 4. Elias Ortega Paredes, Abraham
    et al.
    Nunes, Lauro Roberto
    Dias Santos, Max Mauro
    de Menezes, Leonardo Rodrigues Araujo Xavier
    Silvia Collazos Linares, Kathya
    Francisco Justo, Joao
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Simulation Performance Enhancement in Automotive Embedded Control Using the Unscented Transform2020Inngår i: IEEE Access, E-ISSN 2169-3536, Vol. 8, s. 222041-222049Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Automotive embedded systems comprise several domains, such as in software, electrical, electronics, and control. When designing and testing functions at the top level, one generally ignores the uncertainties arising from the electrical and electronic effects, which could lead to an irregular behavior and deteriorate their performance even using the appropriate methodology for designing the embedded control systems. Then, the studies and comparison on the effect of uncertainty in the automotive domain are important to improve the overall performance of those control systems. Here, we explored the uncertainty in control systems using the Monte Carlo (MC) and unscented transform (UT) methods. These methods have been applied to a mobile seat platform (MSP) and a light emitting diode (LED) used for lighting of heavy-duty vehicles. The UT for embedded control systems has shown better performance when compared to the Monte Carlo method, in order to reduce the number of required variables and computational resources in the simulation of failures and test-case generation. Finally, this investigation brings another application for the UT, in order to exemplify its applicability and advantages when compared with the other methods.

  • 5.
    Feng, Zhiwei
    et al.
    School of Computer Science and Engineering, Northeastern University, Shenyang, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Yu, Haichuan
    School of Computer Science and Engineering, Northeastern University, Shenyang, China.
    Deng, Qingxu
    School of Computer Science and Engineering, Northeastern University, Shenyang, China.
    Niu, Linwei
    Department of Electrical Engineering and Computer Science, Howard University, Washington, USA.
    Online re-routing and re-scheduling of time-triggered flows for fault tolerance in time-sensitive networking2022Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151, Vol. 41, nr 11, s. 4253-4264Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Time-Sensitive Networking (TSN) is an industry-standard networking protocol that is widely deployed in safety-critical industrial and automotive networks thanks to its advantages of deterministic transmission and bounded end-to-end delay for Time-Triggered (TT) flows. In this paper, we focus on TT flows, and address the issue of fault tolerance against permanent and transient faults with both spatial and temporal redundancy. We present an efficient heuristic algorithm for online incremental re-routing and re-scheduling of disrupted flows due to permanent faults, assuming the paths and schedules of existing flows stay fixed and cannot be modified. It is complementary to and can be combined with offline routing and scheduling algorithms for achieving fault tolerance based on Frame Replication and Elimination for Reliability (FRER) (IEEE 802.1CB). Performance evaluation shows that our approach can better recover the system's Degree of Redundancy (DoR) and has a higher acceptance rate than related work.

  • 6.
    Feng, Zhiwei
    et al.
    School of Computer Science and Engineering, Northeastern University, China.
    Wu, Chaoquan
    School of Computer Science and Engineering, Northeastern University, China.
    Deng, Qingxu
    School of Computer Science and Engineering, Northeastern University, China.
    Lin, Yuhan
    School of Computer Science and Engineering, Northeastern University, China.
    Gao, Shichang
    School of Computer Science and Engineering, Northeastern University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    On the scheduling of fault-tolerant time-sensitive networking with IEEE 802.1CB2024Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Time-Sensitive Networking (TSN) has become the most popular technique in modern safety-critical Automotive and Industrial Automation Networks by providing deterministic transmission policies. However, the data of TSN messages may be affected by transient faults. IEEE 802.1CB, a reliability standard in TSN, protects against such faults by providing disjoint redundant routes for each stream. However, the unique assumption may present a new challenge, i.e., an inadequate number of redundant routes that may negatively impact stream scheduling. This paper presents an offline fault-tolerant TSN scheduling approach that considers such impacts for real-time streams (such as Time-Trigger (TT) and Audio Video Bridging (AVB) streams). Specifically, we intend to calculate the minimum upper bound number of disjoint routes required for each stream to meet the reliability requirements, subsequently enhancing the network’s schedulability. We also propose a service degradation function for AVB streams when the network is under heavy load caused by redundant transmissions of TT streams. This function will maintain schedulability and reliability for AVB streams. Experiments with small-and large-scale synthetic networks show the efficiency.

  • 7.
    Gu, Zonghua
    et al.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Jiang, Wei
    University of Electronic Science and Technology of China, China.
    Ho, Tsung-Yi
    The Chinese University of Hong Kong, Hong Kong.
    Preface2022Inngår i: 2022 2nd international conference on intelligent technology and embedded systems (ICITES), IEEE, 2022, s. VI-Kapittel i bok, del av antologi (Annet vitenskapelig)
  • 8.
    He, Qingqiang
    et al.
    The Hong Kong Polytechnic University, Hong Kong, Hong Kong.
    Guan, Nan
    City University of Hong Kong, Hong Kong, Hong Kong.
    Lv, Mingsong
    The Hong Kong Polytechnic University, Hong Kong, Hong Kong; Northeastern University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    On the degree of parallelism in real-time scheduling of DAG tasks2023Inngår i: 2023 Design, Automation & Test in Europe. Conference & Exhibition (DATE): Proceedings, IEEE, 2023, s. 1-6Konferansepaper (Fagfellevurdert)
    Abstract [en]

    Real-time scheduling and analysis of parallel tasks modeled as directed acyclic graphs (DAG) have been intensively studied in recent years. The degree of parallelism of DAG tasks is an important characterization in scheduling. This paper revisits the definition and the computing algorithms for the degree of parallelism of DAG tasks, and clarifies some misunderstandings regarding the degree of parallelism which exist in real-time literature. Based on the degree of the parallelism, we propose a real-time scheduling approach for DAG tasks, which is quite simple but rather effective and outperforms the state-of-the-art by a considerable margin.

  • 9.
    Jiang, Wei
    et al.
    University of Electronic Science and Technology of China, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Ho, Tsung-Yi
    The Chinese University of Hong Kong, Hong Kong.
    Preface2023Inngår i: 2023 IEEE 3rd international conference on intelligent technology and embedded systems (ICITES), Institute of Electrical and Electronics Engineers (IEEE), 2023, s. vi-viKapittel i bok, del av antologi (Fagfellevurdert)
  • 10.
    Jiang, Zhe
    et al.
    University of York, United Kingdom.
    Dai, Xiaotian
    University of York, United Kingdom.
    Burns, Alan
    University of York, United Kingdom.
    Audsley, Neil
    City, University of London, United Kingdom.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gray, Ian
    University of York, United Kingdom.
    A high-resilience imprecise computing architecture for mixed-criticality systems2023Inngår i: IEEE Transactions on Computers, ISSN 0018-9340, E-ISSN 1557-9956, Vol. 72, nr 1, s. 29-42Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Conventional mixed-criticality systems (MCS)s are designed to terminate the execution of less critical tasks in exceptional situations so that the timing properties of more critical tasks can be preserved. Such a strategy can be controversial and has proven difficult to implement in practice, as it can lead to hazards and reduced functionality due to the absence of the discarded tasks. To mitigate this issue, the imprecise mixed-critically system model (IMCS) has been proposed. In such a model, instead of completely dropping less-critical tasks, these tasks are executed as much as possible through the use of decreased computation precision. Although IMCS could effectively improve the survivability of the less-critical tasks, it also introduces three key drawbacks - run-time computation errors, real-time performance degradation, and lack of flexibility. In this paper, we present a novel IMCS framework, which can (i) mitigate the computation errors caused by imprecise computation; (ii) achieve real-time performance near to that of a conventional MCS; (iii) enhance system-level throughput; and (iv) provide flexibility for run-time configuration. We describe the design details of HIART-MCS, and then present the corresponding theoretical analysis and optimisation method for its run-time configuration. Finally, HIART-MCS is evaluated against other MCS frameworks using a variety of experimental metrics.

  • 11.
    Jiang, Zhe
    et al.
    Southeast University, Nanjing, China.
    Dai, Xiaotian
    University of York, United Kingdom.
    Wei, Ran
    University of Cambridge, United Kingdom.
    Gray, Ian
    University of York, United Kingdom.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Zhao, Qingling
    Nanjing University of Science and Technology, China.
    Zhao, Shuai
    Sun Yat-sen University, China.
    NPRC-I/O: a NoC-based real-time I/O system with reduced contention and enhanced predictability2023Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151, s. 1-1Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    All systems rely on inputs and outputs (I/Os) to perceive and interact with their surroundings. In safety-critical systems, it is important to guarantee both the performance and time-predictability of I/O operations. However, with the continued growth of architectural complexity in modern safety-critical systems, satisfying such real-time requirements has become increasingly challenging due to complex I/O transaction paths and extensive hardware contention. In this paper, we present a new NoC-based Predictable I/O system framework (NPRC-I/O) which reduces this contention and ensures the performance and timepredictability of I/O operations. Specifically, NPRC-I/O contains a programmable I/O command controller (NPRC-CC) and a runtime reconfigurable NoC (RNoC), which provides the capability to adjust I/O transaction paths at run-time. Using this flexibility, we construct an end-to-end transmission latency analysis and an optimisation engine that produces configurations for NPRCI/ O and the I/O traffic in a given system. The constructed analysis and optimisation engine guarantee the timing of all hard realtime traffic while reducing the deadline misses of soft real-time traffic and overall transmission latency.

  • 12.
    Li, Liying
    et al.
    School of Computer Science and Engineering, Nanjing University of Science and Technology, China.
    Cong, Peijin
    School of Computer Science and Engineering, Nanjing University of Science and Technology, China.
    Zhou, Junlong
    School of Computer Science and Engineering, Nanjing University of Science and Technology, China; State Key Laboratory of Computer Architecture, Institute of Computing Technology, Chinese Academy of Sciences, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik. Department of Applied Physics and Electronics, Ume˚a University, Sweden.
    Li, Keqin
    Department of Computer Science, State University of New York, United States.
    Data availability optimization for cyber-physical systems2022Inngår i: Proceedings: IEEE Congress on Cybermatics: 2022 IEEE International Conferences on Internet of Things, iThings 2022, IEEE Green Computing and Communications, GreenCom 2022, IEEE Cyber, Physical and Social Computing, CPSCom 2022 and IEEE Smart Data, SmartData 2022, Institute of Electrical and Electronics Engineers (IEEE), 2022, s. 349-356Konferansepaper (Fagfellevurdert)
    Abstract [en]

    As the backbone of Industry 4.0, Cyber-Physical Systems (CPSs) have attracted extensive attention from industry, academia, and government. Missing data is a common problem in CPS data processing and may cause incorrect results and eventually serious malfunction. Existing data availability optimization methods either rely on a large amount of complete training data or suffer from poor performance. To solve these problems, this paper proposes an iterative data availability optimization method for CPSs. Specifically, the proposed method first pre-processes the raw dataset by using a Singular Value Decomposition-based feature selection approach to identify crucial features and reduce computation overheads. It then makes an initial guess for missing values via a designed K-Means-based imputation approach. The appropriate initial estimation decreases the probability of the proposed method falling into the local optimum. Finally, the proposed method iteratively estimates missing data based on the Orthogonal Matching Pursuit algorithm. The proposed method optimizes data availability by accurately imputing missing values. Simulation results on two datasets demonstrate that compared to multiple state-of-the-art approaches, the proposed data availability optimization method can reduce imputation error by up to 99.65%.

  • 13.
    Lin, Wenwei
    et al.
    Sun Yat-Sen University, China.
    Zhong, Chonghao
    Sun Yat-Sen University, China.
    Sun, Xunpei
    Sun Yat-Sen University, China.
    Meng, Haitao
    Technical University of Munich, Germany.
    Chen, Gang
    Sun Yat-Sen University, China.
    Hu, Biao
    China Agricultural University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Rotation-invariant descriptors learned with circulant convolution neural networks2023Inngår i: 2023 IEEE 35th International Conference on Tools with Artificial Intelligence (ICTAI), IEEE, 2023, s. 415-422Konferansepaper (Fagfellevurdert)
    Abstract [en]

    Extracting local features for accurate correspondences between image pairs is an essential basis for various computer vision tasks. Recent works have shown that deep neural networks (DNNs) have demonstrated promising performance in challenging environments. However, these state-of-the-art DNN-based approaches are not well suited for the scenario with geometry rotations due to their intrinsic deficiencies of square kernel structure. That is, square kernel structures in standard DNNs cannot fully identify the essentials of the rotations in geometry. To address this problem, we present RICNN, a novel deep learning framework that encodes invariance against the rotations in geometry explicitly into convolutional neural networks. Rather than using the square-shaped kernel structure, RICNN adopts sectorshaped convolutional kernels to achieve encoding invariance in all rotations. With the explicitness of such rotation encoding, RICNN enables the transfer of perspective DNN models to obtain rotation-invariant descriptions. Furthermore, we propose a novel multi-level hinge triplet loss function to strengthen the matching constraints against geometry rotations. Comprehensive experiments demonstrate the strong generalization ability of the RICNN descriptor on the HPatches dataset. Toward the rotation invariance evaluation, our method shows state-of-the-art results.

  • 14.
    Luan, Siyu
    et al.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Freidovich, Leonid B.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik. Department of Information Technologies and AI, Sirius University of Science and Technology, Sochi, Russian Federation.
    Jiang, Lili
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för datavetenskap.
    Zhao, Qingling
    College of Computer Science and Technology, Nanjing University of Science and Technology, Nanjing, China.
    Out-of-Distribution Detection for Deep Neural Networks with Isolation Forest and Local Outlier Factor2021Inngår i: IEEE Access, E-ISSN 2169-3536, Vol. 9, s. 132980-132989Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Deep Neural Networks (DNNs) are extensively deployed in today's safety-critical autonomous systems thanks to their excellent performance. However, they are known to make mistakes unpredictably, e.g., a DNN may misclassify an object if it is used for perception, or issue unsafe control commands if it is used for planning and control. One common cause for such unpredictable mistakes is Out-of-Distribution (OOD) input samples, i.e., samples that fall outside of the distribution of the training dataset. We present a framework for OOD detection based on outlier detection in one or more hidden layers of a DNN with a runtime monitor based on either Isolation Forest (IF) or Local Outlier Factor (LOF). Performance evaluation indicates that LOF is a promising method in terms of both the Machine Learning metrics of precision, recall, F1 score and accuracy, as well as computational efficiency during testing.

    Fulltekst (pdf)
    fulltext
  • 15.
    Luan, Siyu
    et al.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Saremi, Amin
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Freidovich, Leonid B.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Jiang, Lili
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för datavetenskap.
    Wan, Shaohua
    Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen, China.
    Timing performance benchmarking of out-of-distribution detection algorithms2023Inngår i: Journal of Signal Processing Systems, ISSN 1939-8018, E-ISSN 1939-8115, Vol. 95, nr 12, s. 1355-1370Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    In an open world with a long-tail distribution of input samples, Deep Neural Networks (DNNs) may make unpredictable mistakes for Out-of-Distribution (OOD) inputs at test time, despite high levels of accuracy obtained during model training. OOD detection can be an effective runtime assurance mechanism for safe deployment of machine learning algorithms in safety–critical applications such as medical imaging and autonomous driving. A large number of OOD detection algorithms have been proposed in recent years, with a wide range of performance metrics in terms of accuracy and execution time. For real-time safety–critical applications, e.g., autonomous driving, timing performance is of great importance in addition to accuracy. We perform a comprehensive and systematic benchmark study of multiple OOD detection algorithms in terms of both accuracy and execution time on different hardware platforms, including a powerful workstation and a resource-constrained embedded device, equipped with both CPU and GPU. We also profile and analyze the internal details of each algorithm to identify the performance bottlenecks and potential for GPU acceleration. This paper aims to provide a useful reference for the practical deployment of OOD detection algorithms for real-time safety–critical applications.

    Fulltekst (pdf)
    fulltext
  • 16.
    Luan, Siyu
    et al.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Wan, Shaohua
    Shenzhen Institute for Advanced Study, University of Electronic Science and Technology of China, Shenzhen, China.
    Efficient performance prediction of end-to-end autonomous driving under continuous distribution shifts based on anomaly detection2023Inngår i: Journal of Signal Processing Systems, ISSN 1939-8018, E-ISSN 1939-8115, Vol. 95, nr 12, s. 1455-1468Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    A Deep Neural Network (DNN)’s prediction may be unreliable outside of its training distribution despite high levels of accuracy obtained during model training. The DNN may experience different degrees of accuracy degradation for different levels of distribution shifts, hence it is important to predict its performance (accuracy) under distribution shifts. In this paper, we consider the end-to-end approach to autonomous driving of using a DNN to map from an input image to the control action such as the steering angle. For each input image with possible perturbations that cause distribution shifts, we design a Performance Prediction Module to compute its anomaly score, and use it to predict the DNN’s expected prediction error, i.e., its expected deviation from the ground truth (optimal) control action, which is not available after deployment. If the expected prediction error is too large, then the DNN’s prediction may no longer be trusted, and remedial actions should be taken to ensure safety. We consider different methods for computing the anomaly score for the input image, including using the reconstruction error of an Autoencoder, or applying an Anomaly Detection algorithm to a hidden layer of the DNN. We present performance evaluation of the different methods in terms of both prediction accuracy and execution time on different hardware platforms, in order to provide a useful reference for the designer to choose among the different methods.

    Fulltekst (pdf)
    fulltext
  • 17.
    Luan, Siyu
    et al.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Xu, Rui
    School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China.
    Zhao, Qingling
    School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, China.
    Chen, Gang
    School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China.
    LRP-based network pruning and policy distillation of robust and non-robust DRL agents for embedded systems2023Inngår i: Concurrency and Computation, ISSN 1532-0626, E-ISSN 1532-0634, Vol. 35, nr 19, artikkel-id e7351Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Reinforcement learning (RL) is an effective approach to developing control policies by maximizing the agent's reward. Deep reinforcement learning uses deep neural networks (DNNs) for function approximation in RL, and has achieved tremendous success in recent years. Large DNNs often incur significant memory size and computational overheads, which may impede their deployment into resource-constrained embedded systems. For deployment of a trained RL agent on embedded systems, it is necessary to compress the policy network of the RL agent to improve its memory and computation efficiency. In this article, we perform model compression of the policy network of an RL agent by leveraging the relevance scores computed by layer-wise relevance propagation (LRP), a technique for Explainable AI (XAI), to rank and prune the convolutional filters in the policy network, combined with fine-tuning with policy distillation. Performance evaluation based on several Atari games indicates that our proposed approach is effective in reducing model size and inference time of RL agents. We also consider robust RL agents trained with RADIAL-RL versus standard RL agents, and show that a robust RL agent can achieve better performance (higher average reward) after pruning than a standard RL agent for different attack strengths and pruning rates.

    Fulltekst (pdf)
    fulltext
  • 18.
    Meng, Haitao
    et al.
    Technical University of Munich, Germany.
    Li, Changcai
    Sun Yat-sen University, China.
    Chen, Gang
    Sun Yat-sen University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Knoll, Alois
    Technical University of Munich, Germany.
    ER3D: An Efficient Real-time 3D object detection framework for autonomous driving2023Inngår i: 2023 IEEE 29th International conference on parallel and distributed systems (ICPADS) / [ed] Cristina Ceballos, IEEE Computer Society, 2023, s. 1157-1164Konferansepaper (Fagfellevurdert)
    Abstract [en]

    3D object detection is a vital computer vision task in mobile robotics and autonomous driving. However, most existing methods have exclusively focused on achieving high accuracy, leading to complex and bulky systems that can not be deployed in a real-time manner. In this paper, we propose the ER3D (Efficient and Real-time 3D) object detection framework, which takes stereo images as input and predicts 3D bounding boxes. Instead of using the complex network architecture, we leverage a fast-but-inaccurate method of semi-global matching (SGM) for depth estimation. To eliminate the accuracy degradation in 3D detection caused by inaccurate depth estimation, we introduce decoupled regression head and 3D distance-consistency loU loss to boost the accuracy performance of the 3D detector with a small computing overhead. ER3D achieves both high-precision and real-time performance to enable practical applications of 3D object detection systems on robotic systems. Extensive experiments with the comparison of the state of the arts demonstrate the superior practicability of ER3D, which achieves comparable detection accuracy with significant leadership on inference efficiency.

  • 19.
    Niu, Linwei
    et al.
    Howard University, the Department of Electrical Engineering and Computer Science, DC, Washington, United States.
    Rawat, Danda B.
    Howard University, the Department of Electrical Engineering and Computer Science, DC, Washington, United States.
    Musselwhite, Jonathan
    Howard University, 2400 Sixth Street NW, DC, Washington, United States.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Deng, Qingxu
    Northeastern University, the School of Computer Science and Engineering, Liaoning, Shenyang, China.
    Energy-constrained scheduling for weakly hard real-time systems using standby-sparing2024Inngår i: ACM Transactions on Design Automation of Electronic Systems, ISSN 1084-4309, E-ISSN 1557-7309, Vol. 29, nr 2, artikkel-id 29Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    For real-time embedded systems, QoS (Quality of Service), fault tolerance, and energy budget constraint are among the primary design concerns. In this research, we investigate the problem of energy constrained standby-sparing for both periodic and aperiodic tasks in a weakly hard real-time environment. The standby-sparing systems adopt a primary processor and a spare processor to provide fault tolerance for both permanent and transient faults. For such kind of systems, we firstly propose several novel standby-sparing schemes for the periodic tasks which can ensure the system feasibility under tighter energy budget constraint than the traditional ones. Then based on them integrated approachs for both periodic and aperiodic tasks are proposed to minimize the aperiodic response time whilst achieving better energy and QoS performance under the given energy budget constraint. The evaluation results demonstrated that the proposed techniques significantly outperformed the existing state-of-the-art approaches in terms of feasibility and system performance while ensuring QoS and fault tolerance under the given energy budget constraint.

  • 20.
    Pan, Zhe
    et al.
    Zhejiang University, College of Computer Science and Technology, Hangzhou, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Jiang, Xiaohong
    Zhejiang University, College of Computer Science and Technology, Hangzhou, China.
    Zhu, Guoquan
    Zhejiang Lab, Artificial Intelligence Research Institute, Research Center for Intelligent Computing Systems, Hangzhou, China.
    Ma, De
    Zhejiang University, College of Computer Science and Technology, Hangzhou, China.
    A modular approximation methodology for efficient fixed-point hardware implementation of the sigmoid function2022Inngår i: IEEE Transactions on Industrial Electronics, ISSN 0278-0046, E-ISSN 1557-9948, Vol. 69, nr 10, s. 10694-10703Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    The sigmoid function is a widely used nonlinear activation function in neural networks. In this article, we present a modular approximation methodology for efficient fixed-point hardware implementation of the sigmoid function. Our design consists of three modules: piecewise linear (PWL) approximation as the initial solution, Taylor series approximation of the exponential function, and Newton-Raphson method-based approximation as the final solution. Its modularity enables the designer to flexibly choose the most appropriate approximation method for each module separately. Performance evaluation results indicate that our work strikes an appropriate balance among the objectives of approximation accuracy, hardware resource utilization, and performance.

  • 21.
    Saremi, Amin
    et al.
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik. Department of Electronics, and Connected Services (EACS), Volvo Technology, Göteborg, Sweden.
    Ramkumar, Balaji
    Department of Statistics and Machine Learning, Linköping University, Linköping, Sweden.
    Ghaffari, Ghazaleh
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    An acoustic echo canceller optimized for hands-free speech telecommunication in large vehicle cabins2023Inngår i: EURASIP Journal on Audio, Speech, and Music Processing, ISSN 1687-4714, E-ISSN 1687-4722, Vol. 2023, nr 1, artikkel-id 39Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Acoustic echo cancelation (AEC) is a system identification problem that has been addressed by various techniques and most commonly by normalized least mean square (NLMS) adaptive algorithms. However, performing a successful AEC in large commercial vehicles has proved complicated due to the size and challenging variations in the acoustic characteristics of their cabins. Here, we present a wideband fully linear time domain NLMS algorithm for AEC that is enhanced by a statistical double-talk detector (DTD) and a voice activity detector (VAD). The proposed solution was tested in four main Volvo truck models, with various cabin geometries, using standard Swedish hearing-in-noise (HINT) sentences in the presence and absence of engine noise. The results show that the proposed solution achieves a high echo return loss enhancement (ERLE) of at least 25 dB with a fast convergence time, fulfilling ITU G.168 requirements. The presented solution was particularly developed to provide a practical compromise between accuracy and computational cost to allow its real-time implementation on commercial digital signal processors (DSPs). A real-time implementation of the solution was coded in C on an ARM Cortex M-7 DSP. The algorithmic latency was measured at less than 26 ms for processing each 50-ms buffer indicating the computational feasibility of the proposed solution for real-time implementation on common DSPs and embedded systems with limited computational and memory resources. MATLAB source codes and related audio files are made available online for reference and further development.

    Fulltekst (pdf)
    fulltext
  • 22. Wan, Shaohua
    et al.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik. College of Computer Science, Zhejiang University, Hangzhou 310027, China.
    Ni, Qiang
    Cognitive computing and wireless communications on the edge for healthcare service robots2020Inngår i: Computer Communications, ISSN 0140-3664, E-ISSN 1873-703X, Vol. 149, s. 99-106Artikkel, forskningsoversikt (Fagfellevurdert)
    Abstract [en]

    In recent years, we have witnessed dramatic developments of mobile healthcare robots, which enjoy many advantages over their human counterparts. Previous communication networks for healthcare robots always suffer from high response latency and/or time-consuming computing demands. Robust and high-speed communications and swift processing are critical, sometimes vital in particular in the case of healthcare robots, to the healthcare receivers. As a promising solution, offloading delay-sensitive and communicating-intensive tasks to the robot is expected to improve the services and benefit users. In this paper, we review several state-of-the-art technologies, such as the human-robot interface, environment and user status perceiving, navigation, robust communication and artificial intelligence, of a mobile healthcare robot and discuss in details the customized demands over offloading the computation and communication tasks. According to the intrinsic demands of tasks over the network usage, we categorize abilities of a typical healthcare robot into alternative classes: the edge functionalities and the core functionalities. Many latency-sensitive tasks, such as user interaction, or time-consuming tasks including health receiver status recognition and autonomous moving, can be processed by the robot without frequent communications with data centers. On the other hand, several fundamental abilities, such as radio resource management, mobility management, service provisioning management, need to update the main body with the cutting-edge artificial intelligence. Robustness and safety, in this case, are the primary goals in wireless communications that AI may provide ground-breaking solutions. Based on this partition, this article refers to several state-of-the-art technologies of a mobile healthcare robot and reviews some challenges to be met for its wireless communications.

  • 23. Wan, Shaohua
    et al.
    Qi, Lianyong
    Xu, Xiaolong
    Tong, Chao
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Deep Learning Models for Real-time Human Activity Recognition with Smartphones2020Inngår i: Mobile Networks and Applications, ISSN 1383-469X, E-ISSN 1572-8153, Vol. 25, nr 2, s. 743-755Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    With the widespread application of mobile edge computing (MEC), MEC is serving as a bridge to narrow the gaps between medical staff and patients. Relatedly, MEC is also moving toward supervising individual health in an automatic and intelligent manner. One of the main MEC technologies in healthcare monitoring systems is human activity recognition (HAR). Built-in multifunctional sensors make smartphones a ubiquitous platform for acquiring and analyzing data, thus making it possible for smartphones to perform HAR. The task of recognizing human activity using a smartphone's built-in accelerometer has been well resolved, but in practice, with the multimodal and high-dimensional sensor data, these traditional methods fail to identify complicated and real-time human activities. This paper designs a smartphone inertial accelerometer-based architecture for HAR. When the participants perform typical daily activities, the smartphone collects the sensory data sequence, extracts the high-efficiency features from the original data, and then obtains the user's physical behavior data through multiple three-axis accelerometers. The data are preprocessed by denoising, normalization and segmentation to extract valuable feature vectors. In addition, a real-time human activity classification method based on a convolutional neural network (CNN) is proposed, which uses a CNN for local feature extraction. Finally, CNN, LSTM, BLSTM, MLP and SVM models are utilized on the UCI and Pamap2 datasets. We explore how to train deep learning methods and demonstrate how the proposed method outperforms the others on two large public datasets: UCI and Pamap2.

  • 24.
    Wan, Shaohua
    et al.
    School of Information and Safety Engineering, Zhongnan University of Economics and Law, Wuhan, China.
    Wiśniewski, Remigiusz
    Division of Information Systems and Cybersecurity, University of Zielona Gora, Zielona Góra, Poland.
    Alexandropoulos, George
    Department of Informatics and Telecommunications, National and Kapodistrian University of Athens, Athens, Greece.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Siano, Pierluigi
    Department of Management and Innovation Systems, University of Salerno, Fisciano, Italy.
    Special Issue on Optimization of Cross-layer Collaborative Resource Allocation for Mobile Edge Computing, Caching and Communication2022Inngår i: Computer Communications, ISSN 0140-3664, E-ISSN 1873-703X, Vol. 181, s. 472-473Artikkel i tidsskrift (Annet vitenskapelig)
  • 25.
    Wan, Shaohua
    et al.
    Department of Computer Science and Engineering, Shaoxing University, Shaoxing, China; School of Information and Safety Engineering, Zhongnan University of Economics and Law, Wuhan, China; State Key Laboratory for Novel Software Technology, Nanjing University, Nanjing, China.
    Xu, Xiaolong
    School of Computer and Software, Nanjing University of Information Science and Technology, Nanjing, China.
    Wang, Tian
    College of Computer Science, Huaqiao University, Xiamen, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    An Intelligent Video Analysis Method for Abnormal Event Detection in Intelligent Transportation Systems2021Inngår i: IEEE transactions on intelligent transportation systems (Print), ISSN 1524-9050, E-ISSN 1558-0016, Vol. 22, nr 7, s. 4487-4495, artikkel-id 9190063Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Intelligent transportation systems pervasively deploy thousands of video cameras. Analyzing live video streams from these cameras is of significant importance to public safety. As streaming video is increasing, it becomes infeasible to have human operators sitting in front of hundreds of screens to catch suspicious activities or detect objects of interests in real-time. Actually, with millions of traffic surveillance cameras installed, video retrieval is more vital than ever. To that end, this article proposes a long video event retrieval algorithm based on superframe segmentation. By detecting the motion amplitude of the long video, a large number of redundant frames can be effectively removed from the long video, thereby reducing the number of frames that need to be calculated subsequently. Then, by using a superframe segmentation algorithm based on feature fusion, the remaining long video is divided into several Segments of Interest (SOIs) which include the video events. Finally, the trained semantic model is used to match the answer generated by the text question, and the result with the highest matching value is considered as the video segment corresponding to the question. Experimental results demonstrate that our proposed long video event retrieval and description method which significantly improves the efficiency and accuracy of semantic description, and significantly reduces the retrieval time.

  • 26.
    Wu, Yirui
    et al.
    Hohai University, Jiangsu Province, Nanjing City, China.
    Zhang, Lilai
    Hohai University, Jiangsu Province, Nanjing City, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Lu, Hu
    Jiangsu University, Jiangsu Province, Zhenjiang City, China.
    Wan, Shaohua
    University of Electronic Science and Technology of China, ,Guangdong Province, Shenzhen City, China.
    Edge-AI-driven framework with efficient mobile network design for facial expression recognition2023Inngår i: ACM Transactions on Embedded Computing Systems, ISSN 1539-9087, E-ISSN 1558-3465, Vol. 22, nr 3, artikkel-id 57Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Facial Expression Recognition (FER) in the wild poses significant challenges due to realistic occlusions, illumination, scale, and head pose variations of the facial images. In this article, we propose an Edge-AI-driven framework for FER. On the algorithms aspect, we propose two attention modules, Arbitrary-oriented Spatial Pooling (ASP) and Scalable Frequency Pooling (SFP), for effective feature extraction to improve classification accuracy. On the systems aspect, we propose an edge-cloud joint inference architecture for FER to achieve low-latency inference, consisting of a lightweight backbone network running on the edge device, and two optional attention modules partially offloaded to the cloud. Performance evaluation demonstrates that our approach achieves a good balance between classification accuracy and inference latency.

  • 27.
    Xu, Rui
    et al.
    Nanjing University of Science and Technology, School of Computer Science and Engineering, Nanjing, China.
    Luan, Siyu
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Zhao, Qingling
    Nanjing University of Science and Technology, School of Computer Science and Engineering, Nanjing, China.
    Chen, Gang
    Sun Yat-sen University, School of Computer Science and Engineering, Guangzhou, China.
    LRP-based Policy Pruning and Distillation of Reinforcement Learning Agents for Embedded Systems2022Inngår i: 2022 IEEE 25th International Symposium on Real-Time Distributed Computing, ISORC 2022, Institute of Electrical and Electronics Engineers (IEEE), 2022Konferansepaper (Fagfellevurdert)
    Abstract [en]

    Reinforcement Learning (RL) is an effective approach to developing control policies by maximizing the agent's reward. Deep Reinforcement Learning (DRL) uses Deep Neural Networks (DNNs) for function approximation in RL, and has achieved tremendous success in recent years. Large DNNs often incur significant memory size and computational overheads, which greatly impedes their deployment into resource-constrained embedded systems. For deployment of a trained RL agent on embedded systems, it is necessary to compress the Policy Network of the RL agent to improve its memory and computation efficiency. In this paper, we perform model compression of the Policy Network of an RL agent by leveraging the relevance scores computed by Layer-wise Relevance Propagation (LRP), a technique for Explainable AI (XAI), to rank and prune the convolutional filters in the Policy Network, combined with fine-Tuning with Policy Distillation. Performance evaluation based on several Atari games indicates that our proposed approach is effective in reducing model size and inference time of RL agents.

  • 28.
    Yang, Zhibin
    et al.
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Bao, Yang
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Yang, Yongqiang
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Huang, Zhiqiu
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Bodeveix, Jean-Paul
    IRIT-University of Toulouse, Toulouse, France.
    Filali, Mamoun
    IRIT-University of Toulouse, Toulouse, France.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Exploiting augmented intelligence in the modeling of safety-critical autonomous systems2021Inngår i: Formal Aspects of Computing, ISSN 0934-5043, E-ISSN 1433-299X, Vol. 33, nr 3, s. 343-384Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Machine learning (ML) is used increasingly in safety-critical systems to provide more complex autonomy to make the system to do decisions by itself in uncertain environments. Using ML to learn system features is fundamentally different from manually implementing them in conventional components written in source code. In this paper, we make a first step towards exploring the architecture modeling of safety-critical autonomous systems which are composed of conventional components and ML components, based on natural language requirements. Firstly, augmented intelligence for restricted natural language requirement modeling is proposed. In that, several AI technologies such as natural language processing and clustering are used to recommend candidate terms to the glossary, as well as machine learning is used to predict the category of requirements. The glossary including data dictionary and domain glossary and the category of requirements will be used in the restricted natural language requirement specification method RNLReq, which is equipped with a set of restriction rules and templates to structure and restrict the way how users document requirements. Secondly, automatic generation of SysML architecture models from the RNLReq requirement specifications is presented. Thirdly, the prototype tool is implemented based on Papyrus. Finally, it presents the evaluation of the proposed approach using an industrial autonomous guidance, navigation and control case study.

  • 29.
    Yang, Zhibin
    et al.
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Xing, Linquan
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Xiao, Yingmin
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Zhou, Yong
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Huang, Zhiqiu
    School of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing, China.
    Xue, Lei
    Shanghai Aerospace Electronic Technology Institute, Shanghai, China.
    Model-based reinforcement learning and neural network-based policy compression for spacecraft rendezvous on resource-constrained embedded systems2023Inngår i: IEEE Transactions on Industrial Informatics, ISSN 1551-3203, E-ISSN 1941-0050, Vol. 19, nr 1, s. 1107-1116Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Autonomous spacecraft rendezvous is very challenging in increasingly complex space missions. In this paper, we present our approach Model-Based Reinforcement Learning for Spacecraft Rendezvous Guidance (MBRL4SRG). We build a Markov Decision Process (MDP) model based on the Clohessy-Wiltshire (C-W) equation of spacecraft dynamics, and use dynamic programming to solve it and generate the decision table as the optimal agent policy. Since the onboard computing system of spacecraft is resource-constrained in terms of both memory size and processing speed, we train a Neural Network (NN) as a compact and efficient function approximation to the tabular representation of the decision table. The NN outputs are formally verified using the verification tool Reluval, and the verification results show that the robustness of the NN is maintained. Experimental results indicate that MBRL4SRG achieves lower computational overhead than the conventional PID algorithm, and has higher trustworthiness and better computational efficiency during training than the MFRL algorithms.

  • 30.
    Yin, Lu
    et al.
    School of Computer Science and Engineering, Nanjing University of Science and Technology, 200 Xiaolingwei Street, Nanjing, China.
    Sun, Jin
    School of Computer Science and Engineering, Nanjing University of Science and Technology, 200 Xiaolingwei Street, Nanjing, China.
    Zhou, Junlong
    School of Computer Science and Engineering, Nanjing University of Science and Technology, 200 Xiaolingwei Street, Nanjing, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Li, Keqin
    Department of Computer Science, State University of New York, New Paltz, NY, USA.
    ECFA: an efficient convergent firefly algorithm for solving task scheduling problems in cloud-edge computing2023Inngår i: IEEE Transactions on Services Computing, E-ISSN 1939-1374, s. 1-14Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    In cloud-edge computing paradigms, the integration of edge servers and task offloading mechanisms has posed new challenges to developing task scheduling strategies. This paper proposes an efficient convergent firefly algorithm (ECFA) for scheduling security-critical tasks onto edge servers and the cloud datacenter. The proposed ECFA uses a probability-based mapping operator to convert an individual firefly into a scheduling solution, in order to associate the firefly space with the solution space. Distinct from the standard FA, ECFA employs a low-complexity position update strategy to enhance computational efficiency in solution exploration. In addition, we provide a rigorous theoretical analysis to justify that ECFA owns the capability of converging to the global best individual in the firefly space. Furthermore, we introduce the concept of boundary traps for analyzing firefly movement trajectories, and investigate whether ECFA would fall into boundary traps during the evolutionary procedure under different parameter settings. We create various testing instances to evaluate the performance of ECFA in solving the cloud-edge scheduling problem, demonstrating its superiority over FA-based and other competing metaheuristics. Evaluation results also validate that the parameter range derived from the theoretical analysis can prevent our algorithm from falling into boundary traps.

  • 31. Zhang, Ming
    et al.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Zheng, Nenggan
    Ma, De
    Pan, Gang
    Efficient Spiking Neural Networks With Logarithmic Temporal Coding2020Inngår i: IEEE Access, E-ISSN 2169-3536, Vol. 8, s. 98156-98167Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    A Spiking Neural Network (SNN) can be trained indirectly by first training an Artificial Neural Network (ANN) with the conventional backpropagation algorithm, then converting it into an equivalent SNN. To reduce the computational cost of the resulting SNN as measured by the number of spikes, we present Logarithmic Temporal Coding (LTC), where the number of spikes used to encode an activation grows logarithmically with the activation value; and the accompanying Exponentiate-and-Fire (EF) neuron model, which only involves efficient bit-shift and addition operations. Moreover, we improve the training process of ANN to compensate for approximation errors due to LTC. Experimental results indicate that the resulting SNN achieves competitive performance in terms of classification accuracy at significantly lower computational cost than related work.

    Fulltekst (pdf)
    fulltext
  • 32.
    Zhang, Yi-Wen
    et al.
    College of Computer Science and Technology, Huaqiao University, China.
    Chen, Rong-Kun
    College of Computer Science and Technology, Huaqiao University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Energy-Aware Partitioned Scheduling of Imprecise Mixed-Criticality Systems2023Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151, s. 1-1Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    We consider partitioned scheduling of an Imprecise Mixed-Criticality (IMC) taskset on a uniform multiprocessor platform, with Earliest Deadline First-Virtual Deadline (EDF-VD) as the uniprocessor task scheduling algorithm, and address the optimization problem of finding a feasible task-to-processor assignment and low-criticality (LO) mode processor speed with the objective of minimizing the system’s average energy consumption in LO mode. We propose a task-to-processor assignment algorithm Criticality-Unaware Worst-Fit Decreasing (CU-WFD) algorithm, which allocates tasks with the Worst-Fit Decreasing (WFD) heuristic method based on utilization values at their respective criticality levels. We determine the energy-efficient speed for each processor based on EDF-VD scheduling, and present our algorithm Energy-Efficient Partitioned Scheduling for Imprecise Mixed-Criticality (EEPSIMC) with the CU-WFD heuristic algorithm to minimize system energy consumption. The experimental results show that our proposed algorithm has good performance in terms both schedulability ratio and normalized energy consumption compared to seven comparison baselines.

  • 33.
    Zhang, Yi-Wen
    et al.
    College of Computer Science and Technology, Huaqiao University, China.
    Ma, Jin-Peng
    College of Computer Science and Technology, Huaqiao University, China.
    Zheng, Hui
    College of Computer Science and Technology, Huaqiao University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Criticality-aware EDF scheduling for constrained-deadline imprecise mixed-criticality systems2024Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151, Vol. 43, nr 2, s. 480-491Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    EDF-VD first focuses on the classic mixed-criticality task model in which all low criticality (LO) tasks are abandoned in the high criticality mode, which is an effective dynamic priority scheduling algorithm for mixed-criticality systems. However, it has low schedulability for the imprecise mixed-criticality (IMC) task model with constrained-deadlines, in which LO tasks are provided graceful degradation services instead of being abandoned. In this paper, we study how to improve schedulability for the IMC tasks model. First, we propose a novel criticality-aware EDF scheduling algorithm (CA-EDF) that tries to delay the LO task execution to improve schedulability. Second, we derive sufficient conditions of schedulability for CA-EDF based on the Demand Bound Function. Finally, we evaluate CA-EDF through extensive simulation. The experimental results indicate that CA-EDF can improve the schedulability ratio by about 13.10% compared to the existing algorithms.

  • 34.
    Zhang, Yi-Wen
    et al.
    College of Computer Science and Technology, Huaqiao University, China.
    Zheng, Hui
    College of Computer Science and Technology, Huaqiao University, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    EDF-based energy-efficient semi-clairvoyant scheduling with graceful degradation2024Inngår i: IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, ISSN 0278-0070, E-ISSN 1937-4151, Vol. 43, nr 2, s. 468-479Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Recent works introduce a semi-clairvoyant model, in which the system mode transition is revealed on the arrival of high criticality jobs. To solve the problem of inconsistency between the correctness criterion for mixed-criticality systems (MCS) with a semi-clairvoyant and the actual situation, we study the problem of schedulability and energy in MCS with the semi-clairvoyant model in this paper. First, we propose a new correctness criterion for MCS with semi-clairvoyant and graceful degradation and develop the schedulability test based on Demand Bound Function methods denoted as SCS-GD. Second, we propose an energy-efficient semi-clairvoyant scheduling algorithm based on SCS-GD denoted as EE-SCS-GD. Finally, we conduct an experimental evaluation of SCS-GD and EE-SCS-GD by synthetically generated task sets. The experimental results show that SCS-GD can improve the schedulability ratio by 5.98% compared to existing algorithms while EE-SCS-GD can save 56.17% energy compared to SCS-GD.

  • 35.
    Zhao, Qingling
    et al.
    The PCA Lab, School of Computer Science and Engineering, Nanjing University of Science and Technology, Systems for High-Dimensional Information of Ministry of Education, Jiangsu Key Lab of Image and Video Understanding for Social Security, Jiangsu, Nanjing, China.
    Chen, Mingqiang
    The PCA Lab, School of Computer Science and Engineering, Nanjing University of Science and Technology, Systems for High-Dimensional Information of Ministry of Education, Jiangsu Key Lab of Image and Video Understanding for Social Security, Jiangsu, Nanjing, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Luan, Siyu
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Zeng, Haibo
    Department of Electrical and Computer Engineering, Virginia Tech, VA, Blacksburg, United States.
    Chakrabory, Samarjit
    Department of Computer Science, University of North Carolina, NC, Chapel Hill, United States.
    CAN bus intrusion detection based on auxiliary classifier GAN and out-of-distribution detection2022Inngår i: ACM Transactions on Embedded Computing Systems, ISSN 1539-9087, E-ISSN 1558-3465, Vol. 21, nr 4, artikkel-id 45Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    The Controller Area Network (CAN) is a ubiquitous bus protocol present in the Electrical/Electronic (E/E) systems of almost all vehicles. It is vulnerable to a range of attacks once the attacker gains access to the bus through the vehicle's attack surface. We address the problem of Intrusion Detection on the CAN bus and present a series of methods based on two classifiers trained with Auxiliary Classifier Generative Adversarial Network (ACGAN) to detect and assign fine-grained labels to Known Attacks and also detect the Unknown Attack class in a dataset containing a mixture of (Normal + Known Attacks + Unknown Attack) messages. The most effective method is a cascaded two-stage classification architecture, with the multi-class Auxiliary Classifier in the first stage for classification of Normal and Known Attacks, passing Out-of-Distribution (OOD) samples to the binary Real-Fake Classifier in the second stage for detection of the Unknown Attack class. Performance evaluation demonstrates that our method achieves both high classification accuracy and low runtime overhead, making it suitable for deployment in the resource-constrained in-vehicle environment.

  • 36.
    Zhao, Qingling
    et al.
    Nanjing University of Science and Technology, Nanjing, Jiangsu, China.
    Qu, Mengfei
    Nanjing University of Science and Technology, Nanjing, Jiangsu, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Zeng, Haibo
    Virginia Tech, Blacksburg, United States.
    Minimizing Stack Memory for Partitioned Mixed-criticality Scheduling on Multiprocessor Platforms2022Inngår i: ACM Transactions on Embedded Computing Systems, ISSN 1539-9087, E-ISSN 1558-3465, Vol. 21, nr 2, artikkel-id 20Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    A Mixed-Criticality System (MCS) features the integration of multiple subsystems that are subject to different levels of safety certification on a shared hardware platform. In cost-sensitive application domains such as automotive E/E systems, it is important to reduce application memory footprint, since such a reduction may enable the adoption of a cheaper microprocessor in the family. Preemption Threshold Scheduling (PTS) is a well-known technique for reducing system stack usage. We consider partitioned multiprocessor scheduling, with Preemption Threshold Adaptive Mixed-Criticality (PT-AMC) as the task scheduling algorithm on each processor and address the optimization problem of finding a feasible task-To-processor mapping with minimum total system stack usage on a resource-constrained multi-processor. We present the Extended Maximal Preemption Threshold Assignment Algorithm (EMPTAA), with dual purposes of improving the taskset's schedulability if it is not already schedulable, and minimizing system stack usage of the schedulable taskset. We present efficient heuristic algorithms for finding sub-optimal yet high-quality solutions, including Maximum Utilization Difference based Partitioning (MUDP) and MUDP with Backtrack Mapping (MUDP-BM), as well as a Branch-And-Bound (BnB) algorithm for finding the optimal solution. Performance evaluation with synthetic task sets demonstrates the effectiveness and efficiency of the proposed algorithms.

  • 37.
    Zheng, Wendong
    et al.
    School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China.
    Zhou, Yu
    School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China.
    Chen, Gang
    School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China.
    Gu, Zonghua
    Umeå universitet, Teknisk-naturvetenskapliga fakulteten, Institutionen för tillämpad fysik och elektronik.
    Huang, Kai
    School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China.
    Towards effective training of robust spiking recurrent neural networks under general input noise via provable analysis2023Inngår i: 2023 IEEE/ACM international conference on computer aided design (ICCAD), IEEE, 2023Konferansepaper (Fagfellevurdert)
    Abstract [en]

    Recently, bio-inspired spiking neural networks (SNN) with recurrent structures (SRNN) have received increasingly more attention due to their appealing properties for energy-efficiently solving time-domain classification tasks. SRNN s are often executed in noisy environments on resource-constrained devices which can however greatly compromise its accuracy. Thus, one fundamental question that remains unanswered is whether a formal analysis under the general input noise disturbances can be obtained to guarantee the robustness of SRNNs. Several studies have shown great promises by optimizing the bound over adverse perturbations based on Lipschitz continuity theorem, but most of these theoretical analysis are confined to convolutional neural networks (CNN). In this work, we take a further step towards robust SRNN training via provable robustness analysis over input noise perturbations. We show that it is feasible to establish bound analysis for evaluating noise sensitivity for SRNN by using the relation between the input current and the membrane potential change magnitude across a time window. Inspired by the theoretical analysis, we next propose a targeted penalty term in the objective function for training robust SRNN. Experimental results show that our solution outperforms the more complicated state-of-the-art methods on the commonly tested Fashion MNIST and CIFAR-IO image classification datasets.

1 - 37 of 37
RefereraExporteraLink til resultatlisten
Permanent link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf