检索结果-南通市图书馆

Energy Efficient Speed Scaling and Task Scheduling for Distributed Computing Systems

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2015年第3期24卷 468-473页

作者： HUANG Jiwei lin chuang CHENG Bo State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Department of Computer Science and Technology Tsinghua University

With the rapid development of the performance of computer systems, energy consumption has been increasing dramatically as well. Energy efficiency has been paid much attention, especially in large-scale distributed computing systems. We propose an effective approach for energy reduction, by dynamic speed scaling and task scheduling simultaneously. Markov models for distributed computing systems are proposed, and detailed analyses of the models are provided. Markov decision processes(MDP) are applied for problem formulation, and MDP algorithms for obtaining the optimal solutions are introduced. The efficacy of our approach is further validated by simulation experiments.

关键词： Energy efficiency Speed scaling Task scheduling Markov decision process Distributed computing system

On applying stochastic network calculus

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2013年第6期7卷 924-942页

作者： chuang lin Yiping DENG Yuming JIANG Department of Computer Science and Technology Tsinghua University Beijing 100084 China China Electric Equipment and Systems Engineering Ltd. Beijing 100084 China Department of Telematics Norwegian University of Science and Technology Trondheim NO-7491 Norway

Performance evaluation plays a crucial role in the design of network systems. Many theoretical tools, including queueing theory, effective bandwidth and network calculus, have been proposed to provide modeling mechanisms and resuits. While these theories have been widely adopted for performance evaluation, each has its own limitation. With that network systems have become more complex and harder to describe, where a lot of uncertainty and randomness exists, to make performance evaluation of such systems tractable, some compromise is often necessary and helpful. Stochas- tic network calculus （SNC） is such a theoretical tool. While SNC is a relatively new theory, it is gaining increasing interest and popularity. In the current SNC literature, much attention has been paid on the development of the theory itself. In addition, researchers have also started applying SNC to performance analysis of various types of systems in recent years. The aim of this paper is to provide a tutorial on the new theoretical tool. Specifically, various SNC traffic models and SNC server models are reviewed. The focus is on how to apply SNC, for which, four critical steps are formalized and discussed. In addition, a list of SNC application topics/areas, where there may exist huge research potential, is presented.

关键词： stochastic network calculus bound performance arrival curve service curve

Hierarchical caches in content-centric networks： modeling and analysis

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2015年第6期9卷 846-859页

作者： Zixiao JIA Jiwei HUANG chuang lin Department of Computer Science and Technology Tsinghua University Beijing 100084 China National Computer Network Emergency Response Technical Team/Coordination Center of China Beijing 100029 China

Content-centric network （CCN） is a new Inter- net architecture in which content is treated as the primitive of communication. In CCN, routers are equipped with con- tent stores at the content level, which act as caches for fre- quently requested content. Based on this design, the Internet is available to provide content distribution services without any application-layer support. In addition, as caches are inte- grated into routers, the overall performance of CCN will be deeply affected by the caching efficiency. In this paper, our aim is to gain some insights on how caches should be designed to maintain a high performance in a cost-efficient way. We try to model the two-layer cache hi- erarchy composed of CCN touters using a two-dimensional discrete-time Markov chain, and develop an efficient algo- rithm to calculate the hit ratios of these caches. Simulations validate the accuracy of our modeling method, and convey some meaningful information which can help us better un- derstand the caching mechanism of CCN.

关键词： CCN cache model analysis

Joint study on VMs deployment, assignment and migration in geographically distributed data centers

在线全文

学校读者我要写书评

暂无评论

Frontiers of Computer Science 2016年第3期10卷 559-573页

作者： chuang lin Min YAO Yin LI Department of Computer Science and Technology Tsinghua University Beijing 100084 China

Enterprises build private clouds to provide IT re- sources for geographically distributed subsidiaries or prod- uct divisions. Public cloud providers like Amazon lease their platforms to enterprise users, thus, enterprises can also rent a number of virtual machines （VMs） from their data centers in the service provider networks. Unfortunately, the network cannot always guarantee stable connectivity for their clients to access the VMs or low-latency transfer among data centers. Usually, both latency and bandwidth are in unstable network environment. Being affected by background traffics, the net- work status can be volatile. To reduce the latency uncertainty of client accesses, enterprises should consider the network status when they deploy data centers or rent virtual data cen- ters from cloud providers. In this paper, we first develop a data center deployment and assignment scheme for an enter- prise to meet its users＇ requirements under uncertain network status. To accommodate to the changes of the network status and users＇ demands, a VMs migration-based redeployment scheme is adopted. These two schemes work in a joint way, and lay out a framework to help enterprises make better use of private or public clouds.

关键词： data center deployment VMs migration Min-Max stochastic optimization

Performance Evaluation and Optimization of Hierarchical Routing in SDN Control Plane

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2018年第2期27卷 342-350页

作者： HU Jie lin chuang ZHANG Puheng Department of Computer Science and Technology Tsinghua University

In Software defined networking(SDN), distributed control plane is becoming a promising solution to tackle with scalability and performance. We propose a tree-like hierarchical routing architecture which employs the divide-and-conquer strategy to enhance routing performance on the distributed control plane. To evaluate and optimize the performance of the hierarchical architecture,we apply queuing theory to modeling routing request arrival and processing, and derive an explicit expression of the routing response time. Then we formulate the resource allocation problem in the hierarchical architecture as an optimization problem whose objective is to minimize the response time subject to queue stability and resource constraints. Using the Karush-Kuhn-Tucker(KKT) condition we obtain the optimal resource allocation policy. Based on that, we theoretically prove that the response time under the hierarchical routing architecture is strictly less than that under the Peer to peer(P2 P) architecture. Finally,we perform numerical analyses to confirm the conclusions.

关键词： Software defined networking(SDN) Control plane Routing Performance evaluation Resource allocation

Scalability Analysis of Request Scheduling in Cloud Computing

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Tsinghua Science and Technology 2019年第3期24卷 249-261页

作者： Chao Xue chuang lin Jie Hu Tsinghua National Laboratory for Information Science and Technology (TNList) & Department of Computer Science and Technology Tsinghua University

Rapid advancement of distributed computing systems enables complex services in remote computing clusters. Massive applications with large-scale and disparate characteristics also create high requirements for computing systems. Cloud computing provides a series of novel approaches to meet new trends and ***, some scalability issues have to be addressed in the request scheduling process and few studies have been conducted to solve these problems. Thus, this study investigates the scalability of the request scheduling process in cloud computing. We provide a theoretical definition of the scalability of this process. By modeling the scheduling server as a stochastic preemptive priority queue, we conduct a comprehensive theoretical and numerical analysis of the scalability metric under different structures and various environment configurations. The comparison and conclusion are expected to shed light on the future design and deployment of the request scheduling process in cloud computing.

关键词： cloud computing request scheduling scalability model evaluation

A Review on the Architecture of Software Defined Network

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2018年第6期27卷 1111-1117页

作者： lin chuang HU Jie LI Guoliang CUI Laizhong College of Computer Science and Software Engineering Shenzhen University Department of Computer Science and Technology Tsinghua University

Software defined network(SDN) is a promising network architecture which can simplify network management, ease deployment of network applications, and enable independent evolution of control plane and data plane. However, the separation of control plane and data plane intensifies performance concerns. In this paper, we review the performance of SDN from the perspective of I Ching which has a history of more than two and half millennia of commentary and interpretation, and provides inspiration to the worlds of business, military and religion. We utilize I Ching to model and abstract the architecture of SDN, and discuss the architecture and performance of SDN based on the statements of I Ching. Then we study current works and researches on the performance and architecture of SDN, and discuss the promising directions to the development and evolvement of SDN. Finally,we study common control functions in SDN and the deployment of SDN in different networking settings.

关键词： Software defined networking(SDN) Network architecture I Ching

Design and analysis of an ONOFF variable structure controller for AQM routers supporting TCP flows

在线全文

同方期刊数据库

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 1785-1803页

作者： REN FengYuan YIN YunHe lin chuang Department of Computer Science and Technology Tsinghua University Beijing 100084 China School of Electronics and Information Engineering Beijing Jiaotong University Beijing 100044 China

Active queue management （AQM） can maintain smaller queuing delay and higher throughput by purposefully dropping packets at intermediate nodes. Most of the existing AQM schemes follow the probability dropping mechanism originating from random early detection （RED）. This paper develops a novel packet dropping mechanism for AQM through designing an ONOFF controller applying the variable structure control theory. Because the binary ONOFF controller can considerably simplify the manipulation on the AQM router, it is helpful for implementing the high performance router. The design principles of ONOFF controller are discussed in detail. The guidelines towards parameter settings are presented. The performance is extensively evaluated and compared with other well-known controllers through simulations and theoretical analysis. The results demonstrate that the ONOFF controller is responsive and robust against external disturbances, and is insensitive to variances of the system parameters. Therefore, it is very suitable for the time- varying network system, and at the same time, it can also keep the instantaneous queue length at a desired level with rather small oscillations, which is conducive to achieving the technical objectives of AQM.

关键词： congestion control active queue management ONOFF controller variable structure control

Stochastic Workflow Nets Based Workflow Pattern Modeling

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2014年第1期23卷 6-12页

作者： ZHOU Huan lin chuang MENG Kun CHEN Yarui Tsinghua National Laboratory for Information Science and Technology Department of Computer Science and Technology Tsinghua University

Petri nets are widely used in describing workflow models of management systems, and also functioning in process management and resource *** number of workflow patterns are increasing along with the demands of business process modeling. The Petri net and its extensions can no longer meet the requirement of model description, due to the limitation of their *** models of some patterns are very sophisticated and even grow exponentially in complexity. Also the performance of these new patterns are not well studied. This paper extends Petri nets to Stochastic workflow nets(SWNs)and introduces stochastic or state related variables into arc weights, transition enabling guard and execution *** the new features, SWNs can describe models more accurately and flexibly. And performance of common patterns is calculated under the assumption of exponentially distributed time. The paper also illustrates how to simplify the model with equivalent patterns.

关键词： Stochastic workflow nets Workflow patterns Performance analysis