A5102730151- Graph Theory and Algorithms
- Advanced Graph Neural Networks
- Cloud Computing and Resource Management
- Parallel Computing and Optimization Techniques
- Caching and Content Delivery
- Advanced Sensor and Control Systems
- Plant Pathogens and Fungal Diseases
- Advanced Algorithms and Applications
- Advanced Data Storage Technologies
- Distributed systems and fault tolerance
- Embedded Systems and FPGA Design
- Cloud Data Security Solutions
- Complex Network Analysis Techniques
- Neural Networks and Applications
- Water Quality Monitoring Technologies
- Interconnection Networks and Systems
- Fault Detection and Control Systems
- Advanced Memory and Neural Computing
- Wireless Networks and Protocols
- Underwater Vehicles and Communication Systems
- Face and Expression Recognition
- Distributed and Parallel Computing Systems
- Data Management and Algorithms
- Cloud Computing and Remote Desktop Technologies
- Fungal Biology and Applications
Huazhong University of Science and Technology
2010-2025
Zhejiang Lab
2022-2024
Guizhou University
2009-2024
Peking University
2018-2024
Zhengzhou University
2024
China University of Petroleum, Beijing
2019
Fudan University
2018
Beijing Tian Tan Hospital
2018
Capital Medical University
2018
University of Electronic Science and Technology of China
2018
Computing Power Networking (CPN) represents a transformative paradigm in distributed computing, harnessing the collective capabilities of edge servers dispersed across diverse geographical locations. CPN's core strengths lie its ability to accelerate data processing, diminish latency, and scale efficiently, rendering it particularly apt for real-time applications Internet Things. When coupled with blockchain technology, CPN extends potential by facilitating secure transparent allocation...
Many solutions have been recently proposed to support the processing of streaming graphs. However, for each graph snapshot a graph, new states vertices affected by updates are propagated irregularly along topology. Despite years' research efforts, existing approaches still suffer from serious problems redundant computation overhead and irregular memory access, which severely underutilizes many-core processor. To address these issues, this paper proposes topology-driven programmable...
With the rapidly growing demand of graph processing in real world, a large number iterative jobs run concurrently on same underlying graph. However, storage engines existing frameworks are mainly designed for running an individual job. Our studies show that they inefficient when concurrent due to redundant data and access overhead. To cope with this issue, we develop efficient system, called GraphM. It can be integrated into systems efficiently support higher throughput by fully exploiting...
Dictyophora rubrovolvata is an important edible mushroom that widely cultivated in China. In 2019, a serious rot disease on D. was observed production facility located Ce Heng County, Southwest of Guizhou Province, The causal agent identified as Trichoderma koningiopsis by amplification and sequencing the internal transcribed spacer (ITS) region, translation elongation factor 1-alpha (EF-1α) gene, RNA polymerase II subunit (RPB2) gene followed phylogenetic analysis. Koch's postulates were...
Streaming graph has been broadly employed across various application domains. It involves updating edges to the and then performing analytics on updated graph. However, existing solutions either suffer from poor data locality high computation complexity for streaming analytics, or need overhead search move ensure ordered neighbors during update.
<italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Graph convolutional networks</i> (GCNs) are promising to enable machine learning on graph data. GCNs show potential vertex-level and intra-vertex parallelism for GPU acceleration, but their irregular memory accesses arising in aggregation operations the inherent sparsity vertex features of graphs cause inefficiencies GPU. In this paper, we present gPIM, which aims accelerate inference through a...
In many applications of the analysis dynamic graph, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Timing iterative Graph Processing</i> (TGP) jobs usually need to be generated for processing corresponding snapshots graph obtain results at different points time. For high throughput such applications, it is expected run TGP on GPU concurrently. Although GPU-based systems have been recently developed, out-of-GPU-memory processing, this...
Distributed graph processing platforms usually need to handle massive Concurrent iterative Graph Processing (CGP) jobs for different purposes. However, existing distributed systems face high ratio of data access cost computation the CGP jobs, which incurs low throughput. We observed that there are strong spatial and temporal correlations among accesses issued by because these concurrently running repeatedly traverse shared structure each vertex. Based on this observation, article proposes a...
Many iterative graph processing systems have recently been developed to analyze graphs. Although they are effective from different aspects, there is an important issue that has not addressed yet. A real-world follows the power-law property, in which a small number of vertices high degrees (i.e., connected most other graph). These called <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">hot-vertices</i> and usually require more iterations...
This paper presents the design and implementation of a novel sliding mode control integrated with active disturbance rejection (SMCDR) for precise robust trajectory tracking piezoelectric nanopositioning stages. The model uncertainties, nonlinearity, external disturbances stage are regarded as lumped disturbance, which is estimated by an extended state observer. (ADRC) strategy used to realize preliminary tracking, while adopted handle estimation error residual improve performance....
Higher-order graph clustering aims to partition the using frequently occurring subgraphs (i.e., motifs), instead of lower-order edges, as atomic unit, which has been recognized state-of-the-art solution in ground truth community detection and knowledge discovery. Motif conductance is one most promising higher-order models due its strong interpretability. However, existing motif based algorithms are mainly limited by a seminal two-stage reweighting computing framework, needing enumerate all...
Through a combination of ZigBee technology and infrared remote control technology, air-condition controller can be built to realize functions like monitoring, control, will more intelligent. On the basis protocols development, system uses low-power chip MSP430 CC2530 RF chip, air-condition, monitor adjust indoor temperature humidity, other functions. This paper introduces compiling between CC2530, process data collection uploading. Experimental tests show that operate stably reliably. There...
Graph analytics, which mainly includes graph processing, mining, and learning, has become increasingly important in several domains, including social network analysis, bioinformatics, machine learning. However, analytics applications suffer from poor locality, limited bandwidth, low parallelism owing to the irregular sparse structure, explosive growth, dependencies of data. To address those challenges, programming models, execution modes, messaging strategies are proposed improve utilization...
Dynamic Graph Neural Network (DGNN) has recently attracted a significant amount of research attention from various domains, because most real-world graphs are inherently dynamic. Despite many efforts, for DGNN, existing hardware/software solutions still suffer significantly redundant computation and memory access overhead, they need to irregularly recompute all graph data each snapshot. To address these issues, we propose an efficient redundancy-aware accelerator, RACE , which enables...
Determining the shortest-path distance between vertices in weighted graph is an important problem for a broad range of fields, such as context-aware search and route selection. While many efficient methods querying have been proposed, they are poorly suited parallel architectures, multi-core CPUs or computer clusters, due to strong task dependencies. In this paper, we propose ParaPLL, new parallelism-friendly framework fast query on large-scale graphs. ParaPLL exploits intra-node inter-node...
With the huge demand for graph analytics in many real-world applications, massive iterative processing jobs are concurrently performed on same graphs and suffer from significant high data access cost. To lower cost toward performance, several out-of-core concurrent solutions recently designed to handle by enabling these share accesses of data. However, set active vertices each partition usually different various also evolve with time, where some high-degree ones (or called <italic...
With the increasing need for graph analysis, massive Concurrent iterative Graph Processing (CGP) jobs are usually performed on common large-scale real-world graph. Although several solutions have been proposed, these CGP not coordinated with consideration of inherent dependencies in data driven by topology. As a result, they suffer from redundant and fragmented accesses same underlying dispersed over distributed platform, because is typically irregularly traversed along different paths at...
Streaming Graph Pattern Mining (GPM) has been widely used in many application fields. However, the existing streaming GPM solution suffers from unnecessary explorations and isomorphism tests, while static ones require repetitive operations to compute full graph. In this paper, we propose a pattern-aware incremental execution approach design first accelerator called PSMiner, which integrates multiple optimizations reduce redundant computation improve computing efficiency. We have conducted...
Temporal graph processing is used to handle the snapshots of temporal graph, which concerns changes in over time. Although several software/hardware solutions have been designed for efficient processing, they still suffer from serious irregular data access due uncoordinated traversal. To overcome these limitations, this paper proposes SaGraph, a domain-specific hardware accelerator support graph. Specifically, shows strong similarity, i.e., most accesses different are same and usually refer...