检索结果-内蒙古大学图书馆

IEEE International Conference on Computer Design: VLSI in Computers and Processors, (ICCD)

作者： Zhuohui Duan Haikun Liu Xiaofei Liao Hai Jin Wenbin Jiang Yu Zhang National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab/Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

Non-uniform memory access (NUMA) architectures feature asymmetrical memory access latencies on different CPU nodes. Hybrid memory systems composed of non-volatile memory (NVM) and DRAM further diversify memory access latencies due to the relatively large performance gap between NVM and DRAM. Traditional NUMA memory management policies fail to manage hybrid memories effectively and may even hurt application performance. In this paper, we present HiNUMA, a new NUMA abstraction for memory allocation and migration in hybrid memory systems. HiNUMA advocates NUMA topologyaware hybrid memory allocation policies for the initial data placement. HiNUMA also proposes a new NUMA balancing mechanism called HANB for memory migration at runtime. HANB considers both data access frequency and memory bandwidth utilization to reduce the cost of memory accesses in hybrid memory systems. We evaluate the performance of HiNUMA with several typical workloads. Experimental results show that HiNUMA can effectively utilize hybrid memories, and deliver much higher application performance than conventional NUMA memory management policies and other state-of-the-art work.

关键词： Nonvolatile memory Random access memory Bandwidth Memory management Resource management Task analysis Memory architecture

来源：评论

学校读者我要写书评

暂无评论

FLoc: Fingerprint-Based Indoor Localization system under a Federated Learning Updating Framework

FLoc: Fingerprint-Based Indoor Localization System under a F...

引用

International Conference on Mobile Ad-hoc and Sensor Networks, MSN

作者： Yuxiang Liu Huichuwu Li Jiang Xiao Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9781728152127

ISBN: (纸本)9781728152134

Fingerprinting-based indoor localization via WiFi has achieved a great breakthrough in the past decade. However, it suffers from an inherent problem that the localization accuracy declines sharply over time due to the dynamic environment and unstable WiFi devices. Researchers have designed many methods to update the localization model, e.g., crowdsourcing-based model updating, in order to maintain the localization accuracy. Unfortunately, they have not taken the privacy into consideration during the updating process. This will lead to a threat that the eavesdroppers could guess the location providers' private information according to the updating model. For the goal of maintaining the localization accuracy without the risk of privacy breaching, we proposed FLoc, a fingerprinting-based indoor localization system which updates the localization model via a federated learning framework. In FLoc, every provider maintains a local localization model in their own device. They will regularly encrypt the updating parameters and share them to a common model server. At the model server, it aggregates the encrypted information of the local models to maintain a general model, which will be sent back to the local devices for next updating iteration. We evaluate FLoc in an APs unknown laboratory corridor. The experiment results show that FLoc has a comparable localization performance. Moreover, it can successfully protect the providers' privacy, since the information transferred is all encrypted.

关键词： Mobile handsets Privacy Fingerprint recognition Servers Wireless fidelity Databases Computational modeling

来源：评论

学校读者我要写书评

暂无评论

Minimizing Overheads of Checkpoints in Distributed Stream Processing systems 7

Minimizing Overheads of Checkpoints in Distributed Stream Pr...

引用

7th IEEE International Conference on Cloud Networking, CloudNet 2018

作者： Muhammad Abrar Akber, Syed Chen, Hanhua Wang, Yonghui Jin, Hai School of Computer Science and Technology Huazhong University of Science and Technology Cluster and Grid Computing Lab Services Computing Technology and System Lab Big Data Technology and System Lab Wuhan430074 China

ISBN: (纸本)9781538668313

Failure occurrence in large-scale systems is inevitable, which makes the resilience a key challenge for modern systems. Checkpoints with rollback recovery is a well-known approach to provide fault tolerance in distributed systems. The checkpoint based fault tolerance approach periodically persists the application state to reliable storage, which serves as a recovery point in case of failure. These periodic checkpoints are not inline with the failure rate of the systems as many studies conclude that failure occurrence is not periodic. The optimal size of checkpoint interval is a crucial decision, which directly determines the checkpoint overheads. To minimize the checkpoint overheads, we propose to reduce the number of checkpoints during the application execution. We suggest reducing the number of checkpoints by successively increasing the checkpoint intervals. We consider the failure probability of the underlying infrastructure and iteratively increase the checkpoint intervals. The proposed checkpoint approach tailors the checkpoint initializing based on the failure probability. If failure probability is low, it increases the checkpoint interval, and eventually reduces the total number of checkpoints triggered during application timespan. Reducing the total number of checkpoints during application execution results in decreasing the checkpoint overheads. The experiment results show that the proposed checkpoint policy considerably reduces the checkpoint overheads as compared to periodic checkpoints. © 2018 IEEE.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

High performance RDF updates with triplebit + 13

High performance RDF updates with triplebit +

引用

13th International Conference on Digital Information Management, ICDIM 2018

作者： Yuan, Pingpeng Fan, Lijian Jin, Hai Services Computing Technology and System Lab. Cluster and Grid Computing Lab. School of Computer Science Technology Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)9781538652435

The volume of RDF data continues to grow over the past decade and many known RDF datasets have billions of triples. A grant challenge of managing this huge RDF data is how to access this big RDF data efficiently. A popular approach to addressing the problem is to build a full set of permutations of (S, P, O) indexes. Although this approach has shown to accelerate joins by orders of magnitude, the large space overhead limits the scalability of this approach and makes it heavyweight. In this paper, we present TripleBit +, a fast and compact system for updating RDF data. The design of TripleBit + has two salient features. First, the efficient maintenance strategies of TripleBit + reduces both the overhead to update data and indexes. Second, effective maintenance technologies to handle online updates over RDF repositories are proposed. Our experiments show that TripleBit + outperforms RDF-3X, MonetDB, BitMat on LUBM, UniProt, and BTC 2012 benchmark queries and it offers orders of mangnitude performance improvement for some complex join queries. Our design also yields high task rates as high as 660,000 per second and fast average response time of task which is faster than x-RDF-3X and PostgreSQL. © 2018 IEEE.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Hybrid loss guided densely connected convolutional neural network for Ischemic Stroke Lesion segmentation

Hybrid loss guided densely connected convolutional neural ne...

引用

International Conference for Convergence of technology (I2CT)

作者： Saqib Qamar Hai Jin Ran Zheng Mohd Faizan Services Computing Technology and System Lab Cluster and Grid Computing Lab Big Data Technology and System Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China Department of Electronics Aligarh Muslim University India

ISBN: (数字)9781538680759

ISBN: (纸本)9781538680766

The exact position of dead tissue in ischemic stroke lesions plays an important role to cure the life-threatening condition. However, this issue stays difficult caused by variation of ischemic strokes such as shape and location. Fully convolutional neural networks (FCN) have great potential for semantic image segmentation. Recently, UNet based CNN architectures have secured outstanding performance in the application of medical imaging. The primary problem is data imbalanced in the application of such networks which are very common in medical data application such as lesion applications where non-lesion voxels are higher than lesion voxels. Data imbalance generates low recall and high precision on the prediction of trained networks. Biases towards the particular class are not suitable for many medical works where false positives are less than false negatives. A lot of approaches are developed to tackle this issue including balanced sampling, sample re-weighting, and recently focal loss and similarity loss functions. In this paper, we proposed hyper densely connected network along with hybrid loss functions for ischemic stroke lesions segmentation. The densely connected network exploits the contextual information of multi-modalities, and hybrid loss function uses to overcome the data imbalance. We evaluate our performance on ISLES dataset. Our approach performs well as compared to other existing methods.

关键词： Lesions Image segmentation Magnetic resonance imaging Biomedical imaging Convolutional neural networks Training Computer architecture

来源：评论

学校读者我要写书评

暂无评论

Fast Triangle Counting on GPU

Fast Triangle Counting on GPU

引用

IEEE Conference on High Performance Extreme computing (HPEC)

作者： Chuangyi Gui Long Zheng Pengcheng Yao Xiaofei Liao Hai Jin National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Laboratory Cluster and Grid Computing Laboratory Huazhong University of Science and Technology Wuhan China

Triangle counting is one of the most basic graph applications to solve many real-world problems in a wide variety of domains. Exploring the massive parallelism of the Graphics Processing Unit (GPU) to accelerate the triangle counting is prevail. We identify that the stat-of-the-art GPU-based studies that focus on improving the load balancing still exhibit inherently a large number of random accesses in degrading the performance. In this paper, we design a prefetching scheme that buffers the neighbor list of the processed vertex in advance in the fast shared memory to avoid high latency of random global memory access. Also, we adopt the degree-based graph reordering technique and design a simple heuristic to evenly distribute the workload. Compared to the state-of-the-art HEPC Graph Challenge Champion in the last year, we advance to improve the performance of triangle counting by up to 5.9× speedup with> 109 TEPS on a single GPU for many large real graphs from graph challenge datasets.

关键词： Graphics processing units Prefetching Parallel processing Indexes Acceleration Optimization

来源：评论

学校读者我要写书评

暂无评论

VulDeeLocator: A deep learning-based fine-grained vulnerability detector

arXiv

引用

arXiv 2020年

作者： Li, Zhen Zou, Deqing Xu, Shouhuai Chen, Zhaoxuan Zhu, Yawei Jin, Hai The National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab Big Data Security Engineering Research Center School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan430074 China School of Cyber Security and Computer Hebei University Baoding071002 China The Department of Computer Science University of Colorado Colorado Springs CO80918 United States University of Texas San Antonio United States

Automatically detecting software vulnerabilities is an important problem that has attracted much attention from the academic research community. However, existing vulnerability detectors still cannot achieve the vulnerability detection capability and the locating precision that would warrant their adoption for real-world use. In this paper, we present a vulnerability detector that can simultaneously achieve a high detection capability and a high locating precision, dubbed Vulnerability Deep learning-based Locator (VulDeeLocator). In the course of designing VulDeeLocator, we encounter difficulties including how to accommodate semantic relations between the definitions of types as well as macros and their uses across files, how to accommodate accurate control flows and variable define-use relations, and how to achieve high locating precision. We solve these difficulties by using two innovative ideas: (i) leveraging intermediate code to accommodate extra semantic information, and (ii) using the notion of granularity refinement to pin down locations of vulnerabilities. When applied to 200 files randomly selected from three real-world software products, VulDeeLocator detects 18 confirmed vulnerabilities (i.e., true-positives). Among them, 16 vulnerabilities correspond to known vulnerabilities;the other two are not reported in the National Vulnerability Database (NVD) but have been "silently" patched by the vendor of Libav when releasing newer versions. Copyright © 2020, The Authors. All rights reserved.

关键词： Location

来源：评论

学校读者我要写书评

暂无评论

Modeling Distributed Stream Processing systems Under Heavy Workload

Modeling Distributed Stream Processing Systems Under Heavy W...

引用

International Conference on Cyberworlds

作者： Muhammad Mudassar Qureshi Hanhua Chen Hai Jin Huazhong University of Science and Technology Wuhan China National Engineering Research Center for Big Data Technology and System Cluster and Grid Computing Lab Services Computing Technology and System Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

Big data applications play a significant role in diverse fields. Distributed Stream Processing Engines (DSPEs) are widely used to support real time applications efficiently. Partitioning algorithms are used to partition data streams into multiple nodes to process in parallel to gain efficient performance. Aggregation cost is an important factor when process stateful streaming applications using such partitioning algorithms because it plays an important role on performance when final result is being produced in stateful streaming applications. However, impact of aggregation cost in stream processing is not discussed comprehensively in existing literature. We use performance modeling to identify the importance of aggregation cost when workload is high. We implement performance model on a multi-node cluster to predict the same behavior as on single resource performance model. We demonstrate that stateful streaming applications need more resources as compare to stateless applications when workload is high and both stateful and stateless applications are running in the same DSPE. Experiments results show that a stateful streaming application needs more resources compared to a stateless streaming application when both applications are running on the same DSPE when the workload is high. Further experiment results show that the performance modeling may be helpful to predict maximum workload that can be process on a DSPE and increase in parallelism level is not guaranteed to increase the performance of streaming applications.

关键词： Partitioning algorithms Data models Predictive models Heuristic algorithms Big Data applications Throughput

来源：评论

学校读者我要写书评

暂无评论

A survey on graph processing accelerators: Challenges and opportunities

arXiv

引用

arXiv 2019年

作者： Gui, Chuang-Yi Zheng, Long He, Bing-Sheng Liu, Cheng Chen, Xin-Yu Liao, Xiao-Fei Jin, Hai National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab/Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China School of Computing National University of Singapore 117418 Singapore

Graph is a well known data structure to represent the associated relationships in a variety of applications, e.g., data science and machine learning. Despite a wealth of existing efforts on developing graph processing systems for improving the performance and/or energy efficiency on traditional architectures, dedicated hardware solutions, also referred to as graph processing accelerators, are essential and emerging to provide the benefits significantly beyond those pure software solutions can offer. In this paper, we conduct a systematical survey regarding the design and implementation of graph processing accelerator. Specifically, we review the relevant techniques in three core components toward a graph processing accelerator: preprocessing, parallel graph computation and runtime scheduling. We also examine the benchmarks and results in existing studies for evaluating a graph processing accelerator. Interestingly, we find that there is not an absolute winner for all three aspects in graph acceleration due to the diverse characteristics of graph processing and complexity of hardware configurations. We finially present to discuss several challenges in details, and to further explore the opportunities for the future research. Copyright © 2019, The Authors. All rights reserved.

关键词： Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

CNLoc: Channel state information assisted indoor WLAN localization using nomadic access points 1

引用

15th IFIP International Conference on Network and Parallel computing, NPC 2018

作者： Xiao, Jiang Li, Huichuwu Li, He Jin, Hai Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Department of Information and Electronic Engineering Muroran Institute of Technology MuroranHokkaido Japan

ISBN: (数字)9783030056773

ISBN: (纸本)9783030056766

Wireless local area network (WLAN) based indoor localization is expanding its fast-paced adoption to facilitate a variety of indoor location-based services (ILBS). Unfortunately, the performance of current WLAN localization systems relying on fixed access points (APs) deployment is constrained by the spatial localizability variance (SLV) problem that different locations may exhibit significantly distinct localization resolution. Prior approaches tackle this problem through nomadic APs with favorable mobility to dynamically adjust the network topology. However, the lack of prior knowledge of nomadic AP’s position has been a challenge for location distinction and will lead to prohibitive performance degradation. In this paper, we propose and develop CNLoc, a novel CSI-based (Channel State Information) indoor WLAN localization framework to overcome the location uncertainty of nomadic APs. Our implementation and evaluation show that CNLoc can improve the accuracy with unknown location information of nomadic APs. We also discuss some open issues and new possibilities in future nomadic AP based indoor localization. © IFIP International Federation for Information Processing 2018.

关键词： Channel state information

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：