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： Kam Chi Loong Shihao Han Sishuo Liu Ning Lin Zhongrui Wang Department of Electrical and Electronic Engineering The University of Hong Kong Hong Kong China ACCESS - AI Chip Center for Emerging Smart Systems Hong Kong Science Park Hong Kong China School of Astronautics Harbin Institute of Technology Harbin China School of Microelectronics Southern University of Science and Technology Shenzhen China

ISBN: (数字)9798350380408

ISBN: (纸本)9798350380415

computing-in-memory (CIM) is an emerging computing paradigm, offering noteworthy potential for accelerating neural networks with high parallelism, low latency, and energy efficiency compared to conventional von Neumann architectures. However, existing research has primarily focused on hardware architecture and network co-design for large-scale neural networks, without considering resource constraints. In this study, we aim to develop edge-friendly deep neural networks (DNNs) for accelerators based on resistive random-access memory (RRAM). To achieve this, we propose an edge compilation and resource-constrained RRAM-aware neural architecture search (NAS) framework to search for optimized neural networks meeting specific hardware constraints. Our compilation approach integrates layer partitioning, duplication, and network packing to maximize the utilization of computation units. The resulting network architecture can be optimized for either high accuracy or low latency using a one-shot neural network approach with Pareto optimality achieved through the Non-dominated Sorted Genetic Algorithm ii (NSGA-ii). The compilation of mobile-friendly networks, like Squeezenet and MobilenetV3 small can achieve over 80% of utilization and over 6x speedup compared to ISAAC-like framework with different crossbar resources. The resulting model from NAS optimized for speed achieved 5x-30x speedup. The code for this paper is available at https://***/ArChiiii/rram_nas_comp_pack.

关键词： Computer architecture Parallel processing Network architecture In-memory computing Solids Hardware Common Information Model (computing) Neural architecture search Low latency communication Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

Hierarchically adaptive distributed fault diagnosis in mobile ad hoc networks using clustering

Hierarchically adaptive distributed fault diagnosis in mobil...

引用

International conference on Industrial and Information Systems, ICiiS

作者： Nishi Yadav P.M. Khilar Department of Computer Science & Engineering National Institute of Technology Rourkela India

Ad hoc networking allows portable mobile devices to establish communication path without having any central infrastructure. As there is no centralized infrastructure and the mobile devices are moving randomly, this gives rise to various kinds of problems such as routing and detecting faulty mobile nodes in the network. In this paper, the problem of fault diagnosis in mobile ad hoc networks (MANETs) is considered. In fact, fault-diagnosis becomes important building block to establish dependability in MANET. An important problem in MANET is the distributed system-level diagnosis problem whose purpose is to have each fault-free mobile node to determine the state of all the mobile nodes assuming a MANET composed of N nodes that can be faulty or fault-free. This paper uses a hierarchical clustering approach proposed by authors Durate and Nanya for diagnosing nodes in mobile ad hoc networks (MANETs). The proposed diagnosis algorithm is linearly scalable under the assumption that the mobiles may be: (i) crash faulty due to out of range or physical damage and (ii) value faulty due to sending erroneous messages while operating in the field. The generic parameters such as diagnostic latency and message complexity are used for evaluating the proposed diagnosis algorithm. The result shows that diagnosis latency and message complexity is reduced as compared to non-clustering distributed diagnosis algorithm Forward Heartbeat.

关键词： Mobile computing Mobile communication Ad hoc networks Clustering algorithms Heart beat Testing Computer crashes

来源：评论

学校读者我要写书评

暂无评论

Radio Resource Allocation with Inter-node Interference in Full-Duplex OFDMA Networks

Radio Resource Allocation with Inter-node Interference in Fu...

引用

IEEE International conference on Communications

作者： Changwon Nam Changhee Joo Saewoong Bahk Department of ECE INMC Seoul National University School of ECE UNIST

ISBN: (纸本)9781467364300

In-band wireless full-duplex is a promising technology that enables a wireless node to transmit and receive at the same time on the same frequency spectrum. In OFDMA networks, the full-duplex transmission makes the resource allocation problem more challenging, in particular when user devices are not full-duplex capable. In this paper, we investigate the joint problem of subcarrier assignment and power allocation to maximize the sum-rate performance in full-duplex OFDMA networks. To achieve high throughput in the considered network, we propose to use a practical subcarrier assignment condition which allows a subcarrier to be allocated to a pair of uplink and downlink nodes when its inter-node channel gain is lower than its uplink channel gain. Considering this condition and the inter-node interference, we design three resource allocation algorithms which run for;i) uplink first, ii) downlink first, and iii) uplink and downlink in pair. Through simulation, we evaluate our solutions in comparison with conventional schemes with respect to performance gain.

关键词： Uplink Downlink Resource management Interference Tin Wireless communication Mobile computing

来源：评论

学校读者我要写书评

暂无评论

A Cognitive-Based Ego Network Detection System for Mobile Social Networking

A Cognitive-Based Ego Network Detection System for Mobile So...

引用

Joint IFIP Wireless and Mobile Networking conference (WMNC)

作者： Matteo Mordacchini Andrea Passarella Marco Conti IIT-CNR Pisa Italy

In future generation mobile systems information about social networking structures of users will be fundamental, being a key element for social networking applications, and a crucial contextual information for personalising the behaviour of mobile applications/services. In this paper, we focus specifically on the detection of ego networks. They are networks formed by an individual (ego) and all the other people she has a social relationship with. We propose a completely decentralised algorithm that allows each user's mobile device to identify the structure of its user's ego network. The algorithm monitors social interaction patterns between the ego and its peers. It is completely decentralised and runs at each individual node using local information only, scaling with the network size. It does not disclose social interaction patterns, and it is able to dynamically detect changes in the structure of the ego network, being self-adaptive. The algorithm is based on social cognitive heuristics, i.e. Models about how the human brain groups social relationships, described in the cognitive psychology literature. Therefore, our approach reproduces - in users' personal mobile devices - the cognitive processes used by their human users to understand their ego networks' structure. We test it on real datasets of interactions corresponding to (i) physical contacts and (ii) exchange of information in online social networks. We show that in both cases the detected social structures are remarkably consistent with those described in the social sciences literature. In addition, we study the dynamic behaviour of the algorithm, highlighting how such structures evolve dynamically over time.

关键词： Social network services Mobile communication Mobile computing Heuristic algorithms Mobile handsets Algorithm design and analysis Peer-to-peer computing

来源：评论

学校读者我要写书评

暂无评论

Practical Quantum Communication and Cryptography for WDM Optical Networks

引用

AIP conference Proceedings 2004年第1期734卷 3-11页

作者： Prem Kumar Center for Photonic Communication and Computing Department of Electrical and Computer Engineering Northwestern University Evanston IL 60208‐3118 Department of Physics and Astronomy Northwestern University Evanston IL 60208‐3118

Keeping in mind the ubiquitous standard optical fiber for long‐distance transmission and the widespread availability of efficient active and passive fiber devices, we have been developing telecom‐band resources for practical quantum communication and cryptography in wave‐division‐multiplexed (WDM) optical networks. In this talk I present our recent results on two fronts: i) telecom‐band in‐fiber entanglement generation, storage, and long‐distance distribution and ii) quantum‐noise protected high‐speed data encryption through an optically‐amplified WDM line. Along the first front, with our in‐fiber entanglement source all four Bell states can be readily produced and we have demonstrated violation of Bell’s inequalities by up to 10 standard deviations of measurement uncertainty. With such a source we have demonstrated storage of entanglement for up to 1/8 of a millisecond. Furthermore, when each photon of the entangled pair is propagated in separate 25km‐long standard fibers, high visibility quantum interference is still observed, demonstrating that this system is capable of long‐distance (> 50 km) entanglement distribution. Along the second front, we have implemented a new quantum cryptographic scheme, based on Yuen’s KCQ protocol, in which the inherent quantum noise of coherent states of light is used to perform the cryptographic service of data encryption. In this scheme a legitimate receiver, with use of a short, shared, secret‐key, executes a simple binary decision rule on every transmitted bit. An eavesdropper, on the other hand, who does not possess the secret‐key, is subjected to an irreducible quantum uncertainty in each measurement, even with the use of ideal detectors. We have implemented this scheme to demonstrate quantum‐noise‐protected data encryption at 650 Mbps through a 200 km, in‐line amplified, WDM line. The line simultaneously carried two 10 Gbps standard data channels, 100 GHz on either side of the encrypted channel, which shows that this scheme i

关键词：

来源：评论

学校读者我要写书评

暂无评论

A bandwidth trading marketplace for mobile data offloading

A bandwidth trading marketplace for mobile data offloading

引用

IEEE Annual Joint conference: INFOCOM, IEEE Computer and Communications Societies

作者： Stefano Paris Fabio Martisnon Ilario Filippini Lin Chen Fabio Martignon DIMM University of Bergamo Italy LRI Université Paris-Sud DEI Politecnico di Milano Canada Universite Paris-Sud Orsay ÃŽle-de-France FR

The Radio Access Network (RAN) infrastructure represents the most critical part for capacity planning, which usually accounts for peak traffic conditions. A promising approach to increase the RAN capacity and simultaneously reduce its energy consumption is represented by the opportunistic utilization of third party Wi-Fi access devices. In order to foster the utilization of unexploited Internet connections, we propose a new and open market, where a mobile operator can lease the bandwidth made available by third parties (residential users or private companies) through their access points to increase the network capacity and save large amounts of energy. We formulate the offloading problem as a reverse auction considering the most general case of partial covering of the traffic to be offloaded. We discuss the conditions (i) to offload the maximum amount of data traffic according to the capacity of third party access devices, (ii) to foster the participation of access point owners (individual rationality), and (iii) to prevent market manipulation (incentive compatibility). Finally, we propose a greedy algorithm that solves the offloading problem in polynomial time, even for large-size network scenarios.

关键词： Mobile communication Mobile computing Resource management Wireless communication Bandwidth Greedy algorithms Radio access networks

来源：评论

学校读者我要写书评

暂无评论

A Novel Two-Stage Feature Selection Framework for Efficiently Diagnosing Alzheimer's Disease

A Novel Two-Stage Feature Selection Framework for Efficientl...

引用

Data Science and Information Technology (DSIT), International conference on

作者： Pinya Lu Wenzhao Gao Jie Pan Hongqin Yang Xuemei Ding College of Computer and Cyber Security Fujian Normal University Fuzhou China School of Computing Engineering & Intelligent Systems Ulster University Derry~Londonderry NI UK Xiamen Jingyi Zhikang Technology Company Xiamen China College of Photonic and Electronic Engineering Fujian Normal University Fuzhou China

ISBN: (数字)9798350384093

ISBN: (纸本)9798350384109

Research focus on the trade-off between diagnostic accuracy for Alzheimer's disease (AD) and time management for diagnosis is very limited. This study proposes a novel two-stage feature selection framework integrating variance inflation factor (VIF) and NSGA-ii optimization algorithm to obtain a set of sub-items of cognitive and neuropsychological assessments (CNAs) with high diagnostic accuracy and short managing time. What sets the proposed framework apart is that it 1) mitigates multicollinearity issues caused by multiple types of feature selection algorithms employed on multiple CNAs and 2) has the multi-objective optimization capacity of searching Pareto solutions in a reduced decision space. Crucially, we design a VIF-based fast forward search for feature selection which takes into account the cost of managing time for CNAs. Experimental results demonstrate a significantly shortened managing time of 905 seconds (reduced by two-third), with a slightly improved AUC performance of 0.9113 for classifying cognitively normal controls, mild cognitive impairments, and AD patients based on a set of selected combination of sub-items compared to their involved full CNAs, thereby offering promising prospects for accurate diagnosis and fast assessment in clinical practice.

关键词： Accuracy Costs Data science Aerospace electronics Feature extraction Data mining Alzheimer's disease Information technology Optimization

来源：评论

学校读者我要写书评

暂无评论

Proof of X: Experimental Insights on Blockchain Consensus algorithms in Energy Markets

Proof of X: Experimental Insights on Blockchain Consensus Al...

引用

Innovative Smart Grid Technologies (ISGT)

作者： Travis Machacek Milan Biswal Satyajayant Misra The MITRE Corporation New Mexico State University

ISBN: (数字)9781728188973

ISBN: (纸本)9781728188980

The current centralized model of the electricity market is not efficient in performing distributed energy transactions required for the transactive smart grid. One of the prominent solutions to this issue is to integrate blockchain technologies, which promise transparent, tamper-proof, and secure transaction systems specifically suitable for the decentralized and distributed energy markets. Blockchain has already been shown to successfully operate in a microgrid peer-to-peer (P2P) energy market. The prime determinant of different blockchain implementations is the consensus algorithm they use to reach consensus on which blocks/transactions to accept as valid in a distributed environment. Although different blockchain implementations have been proposed independently for P2P energy market in the microgrid, quantitative experimental analyses and comparison of the consensus algorithms that the different blockchains may use for energy markets, has not been studied. Identifying the right consensus algorithm to use is essential for scalability and operation of the energy market. To this end, we evaluate three popular consensus algorithms: (i) proof of work (PoW), (ii) proof of authority (PoA), and (iii) Istanbul Byzantine fault tolerance (IBFT), running them on a network of nodes set up using a network of docker nodes to form a microgrid energy market. Using a series of double auctions, we assess each algorithm's viability using different metrics, such as time to reach consensus and scalability. The results indicate that PoA is the most efficient and scalable consensus algorithm to hold double auctions in the smart grid. We also identified the minimum hardware specification necessary for devices such as smart meters, which may run these consensus algorithms.

关键词： Scalability Consensus algorithm Microgrids Blockchain Smart meters Peer-to-peer computing Smart grids

来源：评论

学校读者我要写书评

暂无评论

Energy Consumption of Batch and Online Data Stream Learning Models for Smartphone-based Human Activity Recognition

Energy Consumption of Batch and Online Data Stream Learning ...

引用

World conference on Complex Systems

作者： Ilham Amezzane Mohamed El Aroussi Amine Berrazzouk Youssef Fakhri Mohamed Bakhouya LaRIT Lab Faculty of Sciences Ibn Tofail University Kenitra Morocco LERMA Lab Faculty of Computing and Logistics International University of Rabat Sala Aljadida Morocco

Online Smartphone-based Human Activity Recognition (SHAR) systems are increasingly used in different kinds of applications such as in smart homes and health monitoring. In order to allow users to train their own personalized models quickly and without privacy concerns, a SHAR model has to be lightweight due to the limited resources of mobile devices. Therefore, time, memory and energy costs have to be analyzed before integrating the learning process in on-device scenarios. In a previous paper, we presented a comparative study for different machine learning (ML) algorithms, regarding training time, accuracy and memory usage, exploiting two different learning approaches: i) CPU-based vs GPU-based batch learning, and ii) Online data stream learning. In this paper, we complete our study by analyzing power consumption for both approaches. Results show that the same previous best models show also the best power values. The impact that has the CPU on the energy cost is also highlighted.

关键词： Training Data models Computational modeling Sensors Energy consumption Power demand Memory management

来源：评论

学校读者我要写书评

暂无评论

STEROCEN: Simulation and Training Environment for Resource Orchestration in Cloud-Edge Networks

STEROCEN: Simulation and Training Environment for Resource O...

引用

International conference on the Network of the Future (NOF)

作者： Mieszko Ferens Diego Hortelano Ignacio de Miguel Ramón J. Durán Barroso Sokol Kosta Aalborg University Copenhagen Denmark Universidad Rey Juan Carlos Madrid Spain Universidad de Valladolid Valladolid Spain

ISBN: (数字)9798350377767

ISBN: (纸本)9798350377774

Large scale deployment of Internet-of-Things (IoT) devices is projected to grow in the coming years. These devices are expected to be low-cost while supporting applications with growing computational demands. To enable the necessary computations, offloading of computational tasks to Edge and Cloud nodes is a fundamental technology. However, orchestration for such networks is a complex problem which affects both the network design and the decision system. To aid in solving this problem, simulation tools are essential for predicting the performance of networks in different conditions and under different orchestration policies. In this paper, we propose STEROCEN, a Cloud-Edge network resource orchestration simulation and training tool which allows for different configurations of up-to a four-layer network composed of: (i) end-device, (ii) Close Edge, (iii) Far Edge, and (iv) Cloud layers. Our tool collects delay metrics for flexibly defined applications, especially in regard to computation in the network nodes and including uncertainty in processing times. Additionally, the tool only needs the initial configuration and an independently defined orchestrator, allowing for testing of many strategies. As an example, we provide results of testing some Deep Reinforcement Learning (DRL) algorithms using the same training and simulation environment.

关键词： Training Measurement Cloud computing Uncertainty Computational modeling Network architecture Delays Servers Internet of Things Testing

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：