检索结果-内蒙古大学图书馆

A collaborative cloud-edge computing framework in distributed neural network

EURASIP JOURNAL ON WIRELESS COMMUNICATIONS AND networkING 2020年第1期2020卷 1-17页

作者： Xu, Shihao Zhang, Zhenjiang Kadoch, Michel Cheriet, Mohamed Beijing Jiaotong Univ Beijing Municipal Commiss Educ Key Lab Commun & Informat Syst Sch Elect & Informat Engn Beijing Peoples R China Beijing Jiaotong Univ Sch Software Engn Beijing 100044 Peoples R China Ecole Technol Super ETS Montreal PQ H3C 1K3 Canada

The emergence of edge computing provides a new solution to big data processing in the Internet of Things (IoT) environment. By combining edge computing with deep neural network, it can make better use of the advantages of multi-layer architecture of the network. However, the current task offloading and scheduling frameworks for edge computing are not well applicable to neural network training tasks. In this paper, we propose a task model offloading algorithm by considering how to optimally deploy neural network model into the edge nodes. An adaptive task scheduling algorithm is also designed to adaptively optimize the task assignment by using the improved ant colony algorithm. Based on them, a collaborative cloud-edge computing framework is proposed, which can be used in the distributed neural network. Moreover, this framework sets up some mechanisms so that the cloud can collaborate with edge computing in the work. The simulation results show that the framework can reduce time delay and energy consumption, and improve task accuracy.

关键词： Edge computing distributed neural network Resource allocation Task offloading

来源：评论

学校读者我要写书评

暂无评论

A Prototype CVNS distributed neural network Using Synapse-Neuron Modules

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS 2012年第7期59卷 1482-1490页

作者： Khodabandehloo, Golnar Mirhassani, Mitra Ahmadi, Majid Univ Windsor Dept Elect & Comp Engn Windsor ON N9B 3P4 Canada

Continuous valued number system (CVNS) has the fulfilment of being used for building resistive-type distributed neural networks (DNNs). CVNS DNNs benefit form scalability and higher noise immunity compared to the conventional DNNs. In this paper, CVNS is used to design synapse-neuron modules to cope with the environment resolution. These modules can be employed as building blocks of distributed neural networks. A 4-3-2 network is designed and trained to justify the functionality of the proposed structure. This prototype is simulated and laid out in 0.18 mu m CMOS technology.

关键词： Common node CVNS digits distributed neural network resolution of environment sigmoidal neuron synapse-neuron module truncated CVNS

来源：评论

学校读者我要写书评

暂无评论

Construction of an Efficient Divided/distributed neural network Model Using Edge Computing

引用

IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS 2020年第10期E103D卷 2072-2082页

作者： Shingai, Ryuta Hiraga, Yuria Fukuoka, Hisakazu Mitani, Takamasa Nakada, Takashi Nakashima, Yasuhiko Nara Inst Sci & Technol Ikoma 6300192 Japan

Modern deep learning has significantly improved performance and has been used in a wide variety of applications. Since the amount of computation required for the inference process of the neural network is large, it is processed not by the data acquisition location like a surveillance camera but by the server with abundant computing power installed in the data center. Edge computing is getting considerable attention to solve this problem. However, edge computing can provide limited computation resources. Therefore, we assumed a divided/distributed neural network model using both the edge device and the server. By processing part of the convolution layer on edge, the amount of communication becomes smaller than that of the sensor data. In this paper, we have evaluated AlexNet and the other eight models on the distributed environment and estimated FPS values with Wi-Fi, 3G, and 5G communication. To reduce communication costs, we also introduced the compression process before communication. This compression may degrade the object recognition accuracy. As necessary conditions, we set FPS to 30 or faster and object recognition accuracy to 69.7% or higher. This value is determined based on that of an approximation model that binarizes the activation of neural network. We constructed performance and energy models to find the optimal configuration that consumes minimum energy while satisfying the necessary conditions. Through the comprehensive evaluation, we found that the optimal configurations of all nine models. For small models, such as AlexNet, processing entire models in the edge was the best. On the other hand, for huge models, such as VGG16, processing entire models in the server was the best. For medium-size models, the distributed models were good candidates. We confirmed that our model found the most energy efficient configuration while satisfying FPS and accuracy requirements, and the distributed models successfully reduced the energy consumption up to 48.6%, and

关键词： convolutional neural network edge computing distributed neural network video compression

来源：评论

学校读者我要写书评

暂无评论

PATTERN-RECOGNITION BY A distributed neural network - AN INDUSTRIAL APPLICATION

引用

neural networkS 1991年第1期4卷 103-121页

作者： YAO, Y FREEMAN, WJ BURKE, B YANG, Q UNIV CALIF BERKELEY DEPT MOLEC & CELL BIOLDIV NEUROBIOLBERKELEYCA 94720

In this report, a distributed neural network of coupled oscillators is applied to an industrial pattern recognition problem. The network stems from the study of the neurophysiology of the olfactory system. It is shown that the network serves as an associative memory, which possesses chaotic dynamics. The problem addressed is machine recognition of industrial screws, bolts, etc. in simulated real time in accordance with tolerated deviations from manufacturing specifications. After preprocessing, inputs are represented as 1 X 64 binary vectors. We show that our chaotic neural network can accomplish this pattern recognition task better than a standard Bayesian statistical method, a neural binary autoassociator, a three-layer feedforward network under back propagation learning, and our earlier olfactory bulb model that relies on a Hopf bifurcation from equilibrium to limit cycle. The existence of the chaotic dynamics provides the network with its capability to suppress noise and irrelevant information with respect to the recognition task. The collective effectiveness of the "cell-assemblies" and the threshold function of each individual channel enhance the quality of the network as an associative memory. The network classifies an uninterrupted sequence of objects at 200 ms of simulated real time for each object. It reliably distinguishes the unacceptable objects (i.e., 100% correct classification), which is a crucial requirement for this specific application. The effectiveness of the chaotic dynamics may depend on the broad spectrum of the oscillations, which may force classification by spatial rather than temporal characteristics of the operation. Further study of this biologically derived model is needed to determine whether its chaotic dynamics rather than other as yet unidentified attributes is responsible for the superior performance, and if so, how it contributes to that end.

关键词： AUTOASSOCIATOR BACK PROPAGATION CHAOS COLLECTIVE EFFECTIVENESS distributed neural network FEATURE ENHANCEMENT neural network INDUSTRIAL APPLICATION MINIMUM DISTANCE CLASSIFIER NONLINEAR DYNAMICS OLFACTORY SYSTEM PATTERN RECOGNITION

来源：评论

学校读者我要写书评

暂无评论

Analog Circuit Fault Diagnosis Based on distributed neural network

引用

JOURNAL OF COMPUTERS 2010年第11期5卷 1747-1754页

作者： Wang, Long Liu, Yanheng Li, Xiaoguang Guan, Jian Song, Qi Jilin Univ Coll Comp Sci & Technol Changchun 130012 Jilin Peoples R China Liaoning Univ Coll Informat Shenyang 110036 Liaoning Peoples R China

In order to solve the problems caused by large dataset, such as the network scale and the training time, a new method of analog circuit fault diagnosis based on distributed neural network is presented in this paper. The model of distributed neural network is simply introduced. The arithmetic mechanism is depicted in detail and the arithmetic description is given. Simulation results show that this method can improve learning speed and carry out fault diagnosis of analog circuit accurately.

关键词： analog circuit fault diagnosis distributed neural network Hebb rule

来源：评论

学校读者我要写书评

暂无评论

Analysis of combustible gas by distributed neural network

Analysis of combustible gas by distributed neural network

引用

2012 International Conference on Power Electronics Engineering and Computer Technology(PEECT 2012)

作者： Tong Ziyuan Tong Minming Dai Xinlian Faculty of Engineering & Information Technologies The University of Sydney Camperdown/Darlington Campus NSW 2006 Australia Sydney Australia College of Information and Electrical Engineering China University of Mining and Technology

By the effect of electric field on the sensitivity of catalytic sensor, different outputs could be obtained in adjusted electric field. The process of output signal could make the mixedcombustible gases analysis possible with single catalytic sensor. In order to process the outputs, distributed neural network was applied. The mixed gases of CH4, CO, C2H6 were applied in data training in experiment. by this way, The model of information recognition was built. The result proved that the analysis method is feasible. The analysis of gases is very effective in monitoring electrical transformer and dangerous environment.

关键词： gas analysis catalytic sensor distributed neural network

来源：评论

学校读者我要写书评

暂无评论

FLEE: A Hierarchical Federated Learning Framework for distributed Deep neural network over Cloud, Edge, and End Device

引用

ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY 2022年第5期13卷 71-71页

作者： Zhong, Zhengyi Bao, Weidong Wang, Ji Zhu, Xiaomin Zhang, Xiongtao Natl Univ Def Technol Coll Syst Engn 137 Yanwachi St Changsha 410003 Hunan Peoples R China

With the development of smart devices, the computing capabilities of portable end devices such as mobile phones have been greatly enhanced. Meanwhile, traditional cloud computing faces great challenges caused by privacy-leakage and time-delay problems, there is a trend to push models down to edges and end devices. However, due to the limitation of computing resource, it is difficult for end devices to complete complex computing tasks alone. Therefore, this article divides the model into two parts and deploys them on multiple end devices and edges, respectively. Meanwhile, an early exit is set to reduce computing resource overhead, forming a hierarchical distributed architecture. In order to enable the distributed model to continuously evolve by using new data generated by end devices, we comprehensively consider various data distributions on end devices and edges, proposing a hierarchical federated learning framework FLEE, which can realize dynamical updates of models without redeploying them. Through image and sentence classification experiments, we verify that it can improve model performances under all kinds of data distributions, and prove that compared with other frameworks, the models trained by FLEE consume less global computing resource in the inference stage.

关键词： Federated learning early exit of inference distributed neural network

来源：评论

学校读者我要写书评

暂无评论

Missile Trajectory Defense Planning based on distributed neural network

引用

Optik 2022年

作者： Zongjian Cai School of Marxism Guangzhou College of Technology and Business Guangzhou Guangdong 510000 China

In this article, the development status, models and learning algorithms of distributed neural networks are investigated, conventional BP algorithms are analyzed, some improved algorithms are proposed, and the principle and structure of neural network PID controllers are discussed. In addition, a distributed neural network is used to predict the trajectory of the missile and develop an appropriate attack prevention plan. At the same time, the construction of the virtual world will be introduced by constructing the virtual sky and terrain in the simulation process, changing the perspective in the virtual world, and completing the ballistic counterattack plan of the missile. The purpose of the anti-attack plan is to provide technical reserves for the further development of the National Missile Defense Program, so that the Ministry of Defense can update the technology demonstration system of the existing bases, reduce risks and accelerate technological updates. In order to solve the problem of peculiar or drastic changes in the existing trajectory tracking deviation, we introduced a new method that takes the trajectory arc as the independent variable, and built a linear variable parameter system for trajectory tracking on this basis, and calculated a new automatic Tracking deviation under variable.

关键词： distributed neural network Linear variable parameter Missile trajectory Anti-Attack planning

来源：评论

学校读者我要写书评

暂无评论

Application of distributed GA-based RBF neural network in Fire Detection

Application of Distributed GA-based RBF Neural Network in Fi...

引用

4th International Conference on Natural Computation (ICNC 2008)

作者： Wang, Hairong Yang, Weiguo Jiang, Huiling Wang, Yun Armed Police Acad Dept Fire Engn Langfang Hebei Province Peoples R China

ISBN: (纸本)9780769533049

distributed neural network based on RBF network is used to establish the model of fire detection in view of complexity of fire process and multiplicity of fire environment. For guaranteeing network Adaptability, the number of subnets neural network and degree of sample belong to subnet are determined by Fuzzy nearest clustering, RBF network is optimized though a improved genetic algorithm by Adaptive Genetic algorithm is improved by Auto-adaptive crossover probability and the mutation probability that determined the adaptation value. The feasibility of adaptability of the model is proved through training and testing in fire detection model

关键词： distributed neural network RBF network genetic algorithm fire detection

来源：评论

学校读者我要写书评

暂无评论

distributed Deep Convolutional neural network For Smart Camera Image Recognition 2017

Distributed Deep Convolutional Neural Network For Smart Came...

引用

11th International Conference on distributed Smart Cameras (ICDSC)

作者： Castillo, Emmanuel Ayuyao Ahmadinia, Ali Calif State Univ San Marcos Dept Comp Sci 333 S Twin Oaks Valley Rd San Marcos TX 92069 USA

ISBN: (纸本)9781450354875

Smart Cameras brought sight to the idea of cameras as sensors for visual data rather than strictly for taking pictures. One set of applications that leverages this idea are image recognition applications. Unfortunately, a main drawback on these applications is the hardware limitations that they compromise with due to many of its applications residing in the embedded domain. As a matter of fact, a Deep Convolutional neural network (DCNN) is the main general algorithm used in handling the image recognition tasks that can be costly in response time, power consumption and data transfer bandwidth. However, the general architecture of a DCNN is consisted of layers that do specific processing tasks. These distribution of layers can be efficiently organized on a system consisted of Smart Cameras that can improve response time, limit power usage and reduce data transfer bandwidth. A concept is proposed to distribute the layers of a DCNN within a distributed set of Smart Cameras on embedded devices, an edge device and the Cloud that may improve response time, power consumption and data transfer bandwidth.

关键词： distributed neural network Edge Computing

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：