检索结果-南通市图书馆

Learning an identity distinguishable space for large scale face recognition

在线全文

学校读者我要写书评

暂无评论

The Journal of China Universities of Posts and Telecommunications 2018年第1期25卷 54-61页

作者： Yue Ting Wang Hongbo cheng shiduan State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications

Implementing face recognition efficiently to real world large scale dataset presents great challenges to existing approaches. The method in this paper was proposed to learn an identity distinguishable space for large scale face recognition in MSR-Bing image recognition challenge （IRC）. Firstly, a deep convolutional neural network （CNN） was used to optimize a 128 B embedding for large scale face retrieval. The embedding was trained via using triplets of aligned face patches from FaceScrub and CASIA-WebFace datasets. Secondly, the evaluation of MSR-Bing IRC was conducted according to a cross-domain retrieval scheme. The real-time retrieval in this paper was benefited from the K-means clustering performed on the feature space of training data. Furthermore, a large scale similarity learning （LSSL） was applied on the relevant face images for learning a better identity space. A novel method for selecting similar pairs was proposed for LSSL. Compared with many existing networks of face recognition, the proposed model was lightweight and the retrieval method was promising as well.

关键词： face recognition convolutional neural network (CNN) triplet loss inception similarity learning

Energy-Performance Tradeoffs in laaS Cloud with Virtual Machine Scheduling

在线全文

学校读者我要写书评

暂无评论

China Communications 2015年第2期12卷 155-166页

作者： DONG Jiankang WANG Hongbo cheng shiduan State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications

In the cloud data centers,how to map virtual machines(VMs) on physical machines(PMs) to reduce the energy consumption is becoming one of the major issues,and the existing VM scheduling schemes are mostly to reduce energy consumption by optimizing the utilization of physical servers or network ***,the aggressive consolidation of these resources may lead to network performance *** view of this,this paper proposes a two-stage VM scheduling scheme:(1) We propose a static VM placement scheme to minimize the number of activating PMs and network elements to reduce the energy consumption;(2) In the premise of minimizing the migration costs,we propose a dynamic VM migration scheme to minimize the maximum link utilization to improve the network *** scheme makes a tradeoff between energy efficiency and network *** design a new twostage heuristic algorithm for a solution,and the simulations show that our solution achieves good results.

关键词： IaaS cloud virtual machine scheduling network performance energy efficiency

On the feasibility and efficacy of control traffic protection in software-defined networks

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2015年第12期58卷 44-62页

作者： HU YanNan WANG WenDong GONG XiangYang QUE XiRong cheng shiduan State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications

Software Defined Networking(SDN) is an emerging networking paradigm that assumes a logically centralized control plane separated from the data plane. Despite all its advantages, separating the control and data planes introduces new challenges regarding resilient communications between the two. That is, disconnections between switches and their controllers could result in substantial packet loss and performance *** paper addresses this challenge by studying the issue of control traffic protection in SDNs with arbitrary numbers of controllers. Specifically, we propose a control traffic protection scheme that combines both local rerouting and constrained reverse path forwarding protections, through which switches can locally react to failures and redirect the control traffic using standby backup forwarding options. Our goal is then to find a set of primary routes for control traffic, called protected control network, where as many switches as possible can benefit from the proposed protection scheme. We formulate the protected control network problem, prove its NP-hardness, and develop an algorithm that reconciles protectability and performance(e.g., switch-to-control latency). Through extensive simulations based on real topologies, we show that our approach significantly improves protectability of control traffic. The results should help further the process of deploying SDN in real-world networks.

关键词： software-defined network protection control traffic resilience optimization

Chaotic Control of Network Traffic

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

China Communications 2010年第5期7卷 112-119页

作者： Yang Tan Guo Shize Jin Yuehui cheng shiduan School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu 611731 P. R. China State Key Laboratory of Networking and Switching Beijing University of Posts and Telecommunications Beijing 100876 P. R. China Information Security Center Beijing University of Posts and Telecommunications Beijing 100876 P. R. China

This paper presents a chaotic control method on network traffic. By this method, the chaotic network traffic can be controlled to pre-assigned equifibrium point according to chaotic prediction and the Largest Lyapunov Exponent （LLE） of the traffic on congested link is reduced, thereby the probability of traffic burst and network congestion can be reduced. Numerical examples show that this method is effective.

关键词： network traffic chaotic control Lyapunov exponent chaotic prediction

Virtual Machine Scheduling for Improving Energy Efficiency in laaS Cloud

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

China Communications 2014年第3期11卷 1-12页

作者： DONG Jiankang WANG Hongbo LIYangyang cheng shiduan State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing 100876China

In IaaS Cloud,different mapping relationships between virtual machines(VMs) and physical machines(PMs) cause different resource utilization,so how to place VMs on PMs to reduce energy consumption is becoming one of the major concerns for cloud *** existing VM scheduling schemes propose optimize PMs or network resources utilization,but few of them attempt to improve the energy efficiency of these two kinds of resources *** paper proposes a VM scheduling scheme meeting multiple resource constraints,such as the physical server size(CPU,memory,storage,bandwidth,etc.) and network link capacity to reduce both the numbers of active PMs and network elements so as to finally reduce energy *** VM scheduling problem is abstracted as a combination of bin packing problem and quadratic assignment problem,which is also known as a classic combinatorial optimization and NP-hard ***,we design a twostage heuristic algorithm to solve the issue,and the simulations show that our solution outperforms the existing PM- or network-only optimization solutions.

关键词： IaaS cloud virtual machinescheduling energy efficiency bin packingproblem quadratic assignment problem

Active probing based Internet service fault management in uncertain and noisy environment

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science in China(Series F) 2008年第11期51卷 1857-1870页

作者： CHU LingWei ZOU ShiHong cheng shiduan WANG WenDong State Key Lab of Networking and Switching Beijing University of Posts and Telecommunications Beijing100876 China

In Internet service fault management based on active probing, uncertainty and noises will affect service fault management. In order to reduce the impact, challenges of Internet service fault management are analyzed in this paper. Bipartite Bayesian network is chosen to model the dependency relationship between faults and probes, binary symmetric channel is chosen to model noises, and a service fault management approach using active probing is proposed for such an environment. This approach is composed of two phases： fault detection and fault diagnosis. In first phase, we propose a greedy approximation probe selection algorithm （GAPSA）, which selects a minimal set of probes while remaining a high probability of fault detection. In second phase, we propose a fault diagnosis probe selection algorithm （FDPSA）, which selects probes to obtain more system information based on the symptoms observed in previous phase. To deal with dynamic fault set caused by fault recovery mechanism, we propose a hypothesis inference algorithm based on fault persistent time statistic （FPTS）. Simulation results prove the validity and efficiency of our approach.

关键词： service management fault management active probing bipartite Bayesian network binary symmetricchannel

ULM： A user-level model for emotion prediction in social networks

在线全文

学校读者我要写书评

暂无评论

The Journal of China Universities of Posts and Telecommunications 2016年第3期23卷 63-69,88页

作者： Wang Qiyao Li Zhengmin Jin Yuehui cheng shiduan Yang Tan State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing 100876 China National Computer Network Emergency Response Technical Team/Coordination Center of China Beijing 100012 China School of Software Engineering Beijing University of Posts and Telecommunications Beijing 100876 China

Users express their emotions in online social networks （OSNs）. Studying emotions is important for understanding user behaviors. Existing methods in emotion prediction mainly use personal emotion and friend emotion to predict target emotion. In this study, public sentiment is introduced to denote the sentiment of the majority in the network. Public conformity is calculated to measure the degree of a user conforming to the public sentiment. According to the public conformity, users are categorized into three classes： Approvers, independents, and starters. A user-level model for emotion prediction is proposed to predict target emotions of different classes of users, taking into account of the public sentiment, individual sentiment, friend sentiment and pseudo-friend sentiment. Relations between the public conformity and the structure of the network are studied. Experiments conducted on Sina Weibo show that the proposed model could achieve performance improvements to some existing methods in most cases.

关键词： public sentiment conformity user-level model emotion prediction social networks

A cross-layer adaptive transmission scheme over correlated fading channels

在线全文

学校读者我要写书评

暂无评论

Frontiers of Electrical and Electronic Engineering in China 2007年第1期2卷 49-56页

作者： XIAO Junfeng QIU Jing cheng shiduan State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China

Conventional adaptive transmission schemes perform poorly in wireless correlated slow-fading channels.A cross-layer adaptive transmission scheme combined with selective repeat automatic repeat request(SR-ARQ)is *** apply a multi-state Markov system model for analyzing the performance of systems and optimizing the selection of modulation levels and packet sizes in correlated fading channels,which is also described by a finite-state Markov chain.A general closed-form expression of the average throughput for our suggested scheme is *** results show that our adaptive scheme combined with SR-ARQ can obtain good performance in correlated fading channels.

关键词： adaptive modulation(AM) adaptive packet size(APS) automatic repeat request(ARQ) correlated fading channel Markov channel

A self-adaptive key flow routing adjustment algorithm

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Electrical and Electronic Engineering in China 2009年第3期4卷 287-294页

作者： Yujie PEI Hongbo WANG shiduan cheng State Key Laboratory of Networking and Switching Technology Beijing University of Posts and TelecommunicationsBeijing 100876China

Congestion may decrease throughput and transfer efficiency of network,and lead to serious degradation of quality of service(QoS)acquired by end *** routing adjustment methods against congestion mostly consider single-point *** cannot deal with multi-point *** paper presents a probability-based routing adjustment algorithm,which solves the interference problem when several flows are adjusted *** multiple points in the network are being or about to be congested,several key flows are rerouted by this algorithm at the same time to make traffic distribution more reasonable so as to avoid ***,the algorithm in this paper confines the path length of key flows to avoid QoS reduction due to overlength of key *** shows that this method,compared with present algorithms,maintains better load balance and path *** it effectively increases the throughput of the network.

关键词： network management load balancing quality of service(QoS) routing adjustment