- Jilin Zhang;Hui Wu;Weijia Chen;Shaojun Wei;Hong Chen;
Convolutional Neural Networks(CNNs) are widely used in computer vision, natural language processing,and so on, which generally require low power and high efficiency in real applications. Thus, energy efficiency has become a critical indicator of CNN accelerators. Considering that asynchronous circuits have the advantages of low power consumption, high speed, and no clock distribution problems, we design and implement an energy-efficient asynchronous CNN accelerator with a 65 nm Complementary Metal Oxide Semiconductor(CMOS) process. Given the absence of a commercial design tool flow for asynchronous circuits, we develop a novel design flow to implement Click-based asynchronous bundled data circuits efficiently to mask layout with conventional Electronic Design Automation(EDA) tools. We also introduce an adaptive delay matching method and perform accurate static timing analysis for the circuits to ensure correct timing. The accelerator for handwriting recognition network(LeNet-5 model)is implemented. Silicon test results show that the asynchronous accelerator has 30% less power in computing array than the synchronous one and that the energy efficiency of the asynchronous accelerator achieves 1.538 TOPS/W,which is 12% higher than that of the synchronous chip.
2021年05期 v.26 565-573页 [查看摘要][在线阅读][下载 2006K] [下载次数：46 ] |[网刊下载次数：0 ] |[引用频次：4 ] |[阅读次数：0 ] - Zhaoyi Yan;Guangyang Gou;Jie Ren;Fan Wu;Yang Shen;He Tian;Yi Yang;Tian-Ling Ren;
Three main ambipolar compact models for Two-Dimensional(2 D) materials based Field-Effect Transistors(2 D-FETs) are reviewed:(1) Landauer model,(2) 2 D Pao-Sah model, and(3) virtual Source Emission-Diffusion(VSED) model. For the Landauer model, the Gauss quadrature method is applied, and it summarizes all kinds of variants, exhibiting its state-of-art. For the 2 D Pao-Sah model, the aspects of its theoretical fundamentals are rederived, and the electrostatic potentials of electrons and holes are clarified. A brief development history is compiled for the VSED model. In summary, the Landauer model is naturally appropriate for the ballistic transport of short channels, and the 2 D Pao-Sah model is applicable to long-channel devices. By contrast, the VSED model offers a smooth transition between ultimate cases. These three models cover a fairly completed channel length range, which enables researchers to choose the appropriate compact model for their works.
2021年05期 v.26 574-591页 [查看摘要][在线阅读][下载 4221K] [下载次数：50 ] |[网刊下载次数：0 ] |[引用频次：0 ] |[阅读次数：0 ] - Jun Du;Jian Song;Yong Ren;Jintao Wang;
Recently, the fifth-generation(5 G) of wireless networks mainly focuses on the terrestrial applications.However, the well-developed emerging technologies in 5 G are hardly applied to the maritime communications,resulting from the lack of communication infrastructure deployed on the vast ocean, as well as different characteristics of wireless propagation environment over the sea and maritime user distribution. To satisfy the expected plethora of broadband communications and multimedia applications on the ocean, a brand-new maritime information network with a comprehensive coverage capacity in terms of all-hour, all-weather, and all-sea-area has been expected as a revolutionary paradigm to extend the terrestrial capacity of enhanced broadband, massive access, ultra-reliable,and low-latency to the vast ocean. Further considering the limited available resource of maritime communication infrastructure, the convergence of broadband and broadcast/multicast can be regarded as a possible yet practical solution for realizing an efficient and flexible resource configuration with high quality of services. Moreover, according to such multi-functionality and all-coverage maritime information network, the monitoring and sensing of vast ocean area relying on massive Ocean of Things and advanced radar techniques can be also supported. Concerning these issues above, this study proposes a Software Defined Networking(SDN) based Maritime Giant Cellular Network(MagicNet) architecture for broadband and multimedia services. Based on this network, the convergence techniques of broadband and broadcast/multicast, and their supporting for maritime monitoring and marine sensing are also introduced and surveyed.
2021年05期 v.26 592-607页 [查看摘要][在线阅读][下载 1665K] [下载次数：69 ] |[网刊下载次数：0 ] |[引用频次：3 ] |[阅读次数：0 ] - Wenhui Fan;Peiyu Chen;Daiming Shi;Xudong Guo;Li Kou;
With the rapid development of artificial intelligence(AI) technology and its successful application in various fields, modeling and simulation technology, especially multi-agent modeling and simulation(MAMS), of complex systems has rapidly advanced. In this study, we first describe the concept, technical advantages, research steps,and research status of MAMS. Then we review the development status of the hybrid modeling and simulation combining multi-agent and system dynamics, the modeling and simulation of multi-agent reinforcement learning,and the modeling and simulation of large-scale multi-agent. Lastly, we introduce existing MAMS platforms and their comparative studies. This work summarizes the current research situation of MAMS, thus helping scholars understand the systematic technology development of MAMS in the AI era. It also paves the way for further research on MAMS technology.
2021年05期 v.26 608-624页 [查看摘要][在线阅读][下载 329K] [下载次数：312 ] |[网刊下载次数：0 ] |[引用频次：14 ] |[阅读次数：0 ] - Yaping Fu;Yushuang Hou;Zifan Wang;Xinwei Wu;Kaizhou Gao;Ling Wang;
Currently, manufacturing enterprises face increasingly fierce market competition due to the various demands of customers and the rapid development of economic globalization. Hence, they have to extend their production mode into distributed environments and establish multiple factories in various geographical locations.Nowadays, distributed manufacturing systems have been widely adopted in industrial production processes. In recent years, many studies have been done on the modeling and optimization of distributed scheduling problems.This work provides a literature review on distributed scheduling problems in intelligent manufacturing systems. By summarizing and evaluating existing studies on distributed scheduling problems, we analyze the achievements and current research status in this field and discuss ongoing studies. Insights regarding prior works are discussed to uncover future research directions, particularly swarm intelligence and evolutionary algorithms, which are used for managing distributed scheduling problems in manufacturing systems. This work focuses on journal papers discovered using Google Scholar. After reviewing the papers, in this work, we discuss the research trends of distributed scheduling problems and point out some directions for future studies.
2021年05期 v.26 625-645页 [查看摘要][在线阅读][下载 518K] [下载次数：153 ] |[网刊下载次数：0 ] |[引用频次：34 ] |[阅读次数：0 ] - Enda Jiang;Ling Wang;Jingjing Wang;
This paper addresses the Energy-Aware Distributed Hybrid Flow Shop Scheduling Problem with Multiprocessor Tasks(EADHFSPMT) by considering two objectives simultaneously, i.e., makespan and total energy consumption. It consists of three sub-problems, i.e., job assignment between factories, job sequence in each factory,and machine allocation for each job. We present a mixed inter linear programming model and propose a Novel MultiObjective Evolutionary Algorithm based on Decomposition(NMOEA/D). We specially design a decoding scheme according to the characteristics of the EADHFSPMT. To initialize a population with certain diversity, four different rules are utilized. Moreover, a cooperative search is designed to produce new solutions based on different types of relationship between any solution and its neighbors. To enhance the quality of solutions, two local intensification operators are implemented according to the problem characteristics. In addition, a dynamic adjustment strategy for weight vectors is designed to balance the diversity and convergence, which can adaptively modify weight vectors according to the distribution of the non-dominated front. Extensive computational experiments are carried out by using a number of benchmark instances, which demonstrate the effectiveness of the above special designs. The statistical comparisons to the existing algorithms also verify the superior performances of the NMOEA/D.
2021年05期 v.26 646-663页 [查看摘要][在线阅读][下载 1997K] [下载次数：81 ] |[网刊下载次数：0 ] |[引用频次：25 ] |[阅读次数：0 ] - Youhui Zhang;Peng Qu;Weimin Zheng;
Brain-inspired computing refers to computational models, methods, and systems, that are mainly inspired by the processing mode or structure of brain. A recent study proposed the concept of "neuromorphic completeness"and the corresponding system hierarchy, which is helpful to determine the capability boundary of brain-inspired computing system and to judge whether hardware and software of brain-inspired computing are compatible with each other. As a position paper, this article analyzes the existing brain-inspired chips design characteristics and the current so-called "general purpose" application development frameworks for brain-inspired computing, as well as introduces the background and the potential of this proposal. Further, some key features of this concept are presented through the comparison with the Turing completeness and approximate computation, and the analyses of the relationship with "general-purpose" brain-inspired computing systems(it means that computing systems can support all computable applications). In the end, a promising technical approach to realize such computing systems is introduced, as well as the on-going research and the work foundation. We believe that this work is conducive to the design of extensible neuromorphic complete hardware-primitives and the corresponding chips. On this basis, it is expected to gradually realize "general purpose" brain-inspired computing system, in order to take into account the functionality completeness and application efficiency.
2021年05期 v.26 664-673页 [查看摘要][在线阅读][下载 496K] [下载次数：40 ] |[网刊下载次数：0 ] |[引用频次：3 ] |[阅读次数：0 ] - Kai Zhu;Tao Zhang;
Navigation is a fundamental problem of mobile robots, for which Deep Reinforcement Learning(DRL)has received significant attention because of its strong representation and experience learning abilities. There is a growing trend of applying DRL to mobile robot navigation. In this paper, we review DRL methods and DRL-based navigation frameworks. Then we systematically compare and analyze the relationship and differences between four typical application scenarios: local obstacle avoidance, indoor navigation, multi-robot navigation, and social navigation. Next, we describe the development of DRL-based navigation. Last, we discuss the challenges and some possible solutions regarding DRL-based navigation.
2021年05期 v.26 674-691页 [查看摘要][在线阅读][下载 4580K] [下载次数：270 ] |[网刊下载次数：0 ] |[引用频次：121 ] |[阅读次数：0 ] - Zhipeng Song;Zhichao Cao;Zhenjiang Li;Jiliang Wang;Yunhao Liu;
Motion tracking via Inertial Measurement Units(IMUs) on mobile and wearable devices has attracted significant interest in recent years. High-accuracy IMU-tracking can be applied in various applications, such as indoor navigation, gesture recognition, text input, etc. Many efforts have been devoted to improving IMU-based motion tracking in the last two decades, from early calibration techniques on ships or airplanes, to recent arm motion models used on wearable smart devices. In this paper, we present a comprehensive survey on IMU-tracking techniques on mobile and wearable devices. We also reveal the key challenges in IMU-based motion tracking on mobile and wearable devices and possible directions to address these challenges.
2021年05期 v.26 692-705页 [查看摘要][在线阅读][下载 465K] [下载次数：47 ] |[网刊下载次数：0 ] |[引用频次：5 ] |[阅读次数：0 ]
- Baohua Sun;Richard Al-Bayaty;Qiuyuan Huang;Dapeng Wu;
Graph clustering, i.e., partitioning nodes or data points into non-overlapping clusters, can be beneficial in a large varieties of computer vision and machine learning applications. However, main graph clustering schemes, such as spectral clustering, cannot be applied to a large network due to prohibitive computational complexity required.While there exist methods applicable to large networks, these methods do not offer convincing comparisons against known ground truth. For the first time, this work conducts clustering algorithm performance evaluations on large networks(consisting of one million nodes) with ground truth information. Ideas and concepts from game theory are applied towards graph clustering to formulate a new proposed algorithm, Game Theoretical Approach for Clustering(GTAC). This theoretical framework is shown to be a generalization of both the Label Propagation and Louvain methods, offering an additional means of derivation and analysis. GTAC introduces a tuning parameter which allows variable algorithm performance in accordance with application needs. Experimentation shows that these GTAC algorithms offer scalability and tunability towards big data applications.
2021年05期 v.26 706-723页 [查看摘要][在线阅读][下载 5948K] [下载次数：51 ] |[网刊下载次数：0 ] |[引用频次：1 ] |[阅读次数：0 ] - Naijin Chen;Zhen Wang;Ruixiang He;Jianhui Jiang;Fei Cheng;Chenghao Han;
Row Parallel Coarse-Grained Reconfigurable Architecture(RPCGRA) has the advantages of maximum parallelism and programmable flexibility. Designing an efficient algorithm to map the diverse applications onto RPCGRA is difficult due to a number of RPCGRA hardware constraints. To solve this problem, the nodes of the data flow graph must be partitioned and scheduled onto the RPCGRA. In this paper, we present a Depth-First Greedy Mapping(DFGM) algorithm that simultaneously considers the communication costs and the use times of the Reconfigurable Cell Array(RCA). Compared with level breadth mapping, the performance of DFGM is better. The percentage of maximum improvement in the use times of RCA is 33% and the percentage of maximum improvement in non-original input and output times is 64.4%(Given Discrete Cosine Transfor 8(DCT8), and the area of reconfigurable processing unit is 56). Compared with level-based depth mapping, DFGM also obtains the lowest averages of use times of RCA, non-original input and output times, and the reconfigurable time.
2021年05期 v.26 724-735页 [查看摘要][在线阅读][下载 822K] [下载次数：40 ] |[网刊下载次数：0 ] |[引用频次：11 ] |[阅读次数：0 ] - Ling Zhang;Jianchao Liu;Fangxing Shang;Gang Li;Juming Zhao;Yueqin Zhang;
Level-set-based image segmentation has been widely used in unsupervised segmentation tasks.Researchers have recently alleviated the influence of image noise on segmentation results by introducing global or local statistics into existing models. Most existing methods are based on the assumption that the distribution of image noise is known or observable. However, real-time images do not meet this assumption. To bridge this gap, we propose a novel level-set-based segmentation method with an unsupervised denoising mechanism. First,a denoising filter is acquired under the unsupervised learning paradigm. Second, the denoising filter is integrated into the level-set framework to separate noise from the noisy image input. Finally, the level-set energy function is minimized to acquire segmentation contours. Extensive experiments demonstrate the robustness and effectiveness of the proposed method when applied to noisy images.
2021年05期 v.26 736-748页 [查看摘要][在线阅读][下载 2779K] [下载次数：43 ] |[网刊下载次数：0 ] |[引用频次：7 ] |[阅读次数：0 ] - Guanjie Liu;Yan Wei;Yunshen Xie;Jianqiang Li;Liyan Qiao;Ji-jiang Yang;
The current mode of clinical aided diagnosis of Ocular Myasthenia Gravis(OMG) is time-consuming and laborious, and it lacks quantitative standards. An aided diagnostic system for OMG is proposed to solve this problem.The values calculated by the system include three clinical indicators: eyelid distance, sclera distance, and palpebra superior fatigability test time. For the first two indicators, the semantic segmentation method was used to extract the pathological features of the patient's eye image and a semantic segmentation model was constructed. The patient eye image was divided into three regions: iris, sclera, and background. The indicators were calculated based on the position of the pixels in the segmentation mask. For the last indicator, a calculation method based on the Eyelid Aspect Ratio(EAR) is proposed; this method can better reflect the change of eyelid distance over time. The system was evaluated based on the collected patient data. The results show that the segmentation model achieves a mean Intersection-Over-Union(mIoU) value of 86.05%. The paired-sample T-test was used to compare the results obtained by the system and doctors, and the p values were all greater than 0.05. Thus, the system can reduce the cost of clinical diagnosis and has high application value.
2021年05期 v.26 749-758页 [查看摘要][在线阅读][下载 2367K] [下载次数：36 ] |[网刊下载次数：0 ] |[引用频次：3 ] |[阅读次数：0 ] - Junjie Pang;Yan Huang;Zhenzhen Xie;Jianbo Li;Zhipeng Cai;
The novel coronavirus, COVID-19, has caused a crisis that affects all segments of the population. As the knowledge and understanding of COVID-19 evolve, an appropriate response plan for this pandemic is considered one of the most effective methods for controlling the spread of the virus. Recent studies indicate that a city Digital Twin(DT) is beneficial for tackling this health crisis, because it can construct a virtual replica to simulate factors,such as climate conditions, response policies, and people's trajectories, to help plan efficient and inclusive decisions.However, a city DTsystem relies on long-term and high-quality data collection to make appropriate decisions, limiting its advantages when facing urgent crises, such as the COVID-19 pandemic. Federated Learning(FL), in which all clients can learn a shared model while retaining all training data locally, emerges as a promising solution for accumulating the insights from multiple data sources efficiently. Furthermore, the enhanced privacy protection settings removing the privacy barriers lie in this collaboration. In this work, we propose a framework that fused city DT with FL to achieve a novel collaborative paradigm that allows multiple city DTs to share the local strategy and status quickly. In particular, an FL central server manages the local updates of multiple collaborators(city DTs),providing a global model that is trained in multiple iterations at different city DT systems until the model gains the correlations between various response plans and infection trends. This approach means a collaborative city DT paradigm fused with FL techniques can obtain knowledge and patterns from multiple DTs and eventually establish a"global view" of city crisis management. Meanwhile, it also helps improve each city's DT by consolidating other DT's data without violating privacy rules. In this paper, we use the COVID-19 pandemic as the use case of the proposed framework. The experimental results on a real dataset with various response plans validate our proposed solution and demonstrate its superior performance.
2021年05期 v.26 759-771页 [查看摘要][在线阅读][下载 12951K] [下载次数：80 ] |[网刊下载次数：0 ] |[引用频次：5 ] |[阅读次数：0 ] - Yu Tian;Ruiqing Zheng;Zhenlan Liang;Suning Li;Fang-Xiang Wu;Min Li;
Recently, the emergence of single-cell RNA-sequencing(scRNA-seq) technology makes it possible to solve biological problems at the single-cell resolution. One of the critical steps in cellular heterogeneity analysis is the cell type identification. Diverse scRNA-seq clustering methods have been proposed to partition cells into clusters.Among all the methods, hierarchical clustering and spectral clustering are the most popular approaches in the downstream clustering analysis with different preprocessing strategies such as similarity learning, dropout imputation,and dimensionality reduction. In this study, we carry out a comprehensive analysis by combining different strategies with these two categories of clustering methods on scRNA-seq datasets under different biological conditions. The analysis results show that the methods with spectral clustering tend to perform better on datasets with continuous shapes in two-dimension, while those with hierarchical clustering achieve better results on datasets with obvious boundaries between clusters in two-dimension. Motivated by this finding, a new strategy, called QRS, is developed to quantitatively evaluate the latent representative shape of a dataset to distinguish whether it has clear boundaries or not. Finally, a data-driven clustering recommendation method, called DDCR, is proposed to recommend hierarchical clustering or spectral clustering for scRNA-seq data. We perform DDCR on two typical single cell clustering methods,SC3 and RAFSIL, and the results show that DDCR recommends a more suitable downstream clustering method for different scRNA-seq datasets and obtains more robust and accurate results.
2021年05期 v.26 772-789页 [查看摘要][在线阅读][下载 2297K] [下载次数：47 ] |[网刊下载次数：0 ] |[引用频次：4 ] |[阅读次数：0 ] 下载本期数据