By Topic

Large Scale Network-Centric Distributed Systems

Cover Image Copyright Year: 2014
Author(s): Sarbazi-Azad, H.; Zomaya, A.
Publisher: Wiley-IEEE Press
Content Type : Books & eBooks
Topics: Communication, Networking & Broadcasting ;  Computing & Processing (Hardware/Software)
  • Print

Abstract

A highly accessible reference offering a broad range of topics and insights on large scale network-centric distributed systemsEvolving from the fields of high-performance computing and networking, large scale network-centric distributed systems continues to grow as one of the most important topics in computing and communication and many interdisciplinary areas. Dealing with both wired and wireless networks, this book focuses on the design and performance issues of such systems.Large Scale Network-Centric Distributed Systems provides in-depth coverage ranging from ground-level hardware issues (such as buffer organization, router delay, and flow control) to the high-level issues immediately concerning application or system users (including parallel programming, middleware, and OS support for such computing systems). Arranged in five parts, it explains and analyzes complex topics to an unprecedented degree:
Part 1: Multicore and Many-Core (Mc) Systems-on-Chip
Part 2: Pervasive/Ubiquitous Computing and Peer-to-Peer Systems
Part 3: Wireless/Mobile Networks
Part 4: Grid and Cloud Computing
Part 5: Other Topics Related to Network-Centric Computing and Its ApplicationsLarge Scale Network-Centric Distributed Systems is an incredibly useful resource for practitioners, postgraduate students, postdocs, and researchers.

  •   Click to expandTable of Contents

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Front Matter

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.fmatter
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      The prelims comprise:
      Half-Title Page
      Wiley Series Page
      Title Page
      Copyright Page
      Dedication Page
      Contents
      Preface
      Acknowledgments
      List of Figures
      List of Tables
      List of Contributors View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      A Reconfigurable On-Chip Interconnection Network for Large Multicore Systems

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch01
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter introduces the baseline reconfigurable networks-on-chip (NoC) architecture. It addresses two problems of core to network mapping and topology exploration in which the cores of a given set of input applications are physically mapped to the network and then a suitable topology is found for each individual application. Experimental results, using some multicore system-on-chips (SoC) workloads, show that this architecture effectively improves the performance of NoCs by 29% and reduces the power consumption by 9% over one of the most efficient and popular mapping algorithms proposed for conventional NoCs. The chapter extends the baseline reconfigurable NoC to a generalized reconfigurable NoC architecture. This new cluster-based structure consists of several mesh clusters alongside a reconfigurable connection fabric that handles the intercluster communication. It can support both local and global traffic patterns in an efficient manner. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Compilers, Techniques, and Tools for Supporting Programming Heterogeneous Many/Multicore Systems

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch02
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter deals with different programming models for current many/multicore and graphical processing units (GPUs) systems. It reviews the state of the art of compilers and support tools tailored to help programmers both in developing and porting code for many/multicore CPUs and GPUs. The chapter focuses on automatic parallelization compilers, techniques, and tools that do source-to-source transformation of code to convert sequential code into parallel. It presents Code Analyser and knowLedge ModEller for Transformation (CALuMET), a prototype tool implementing a component-based architecture designed to assist the user in the process of software migration by showing extracted knowledge from code and by driving automatic transformation of parallelizable regions of code. The chapter integrates an algorithmic recognizer to find instances of known algorithms in the code, model the extracted knowledge, and drive a source-to-source transformer to convert sequential code to a parallel version by using libraries for GPUs. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      A Multithreaded Branch-and-Bound Algorithm for Solving The Flow-Shop Problem on A Multicore Environment

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch03
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      Many real-world problems encountered in different industrial and economic fields, such as task allocation, job scheduling, network routing, cutting, and packing, are of a combinatorial nature. One of the most popular exact methods for solving to optimality a combinatorial optimization problem is the branch-and-bound (B&B) algorithm. This chapter proposes a multithreaded B&B algorithm for solving the flow-shop scheduling problem on a multicore system. It explains, in general, the different models used for parallelization of a B&B. The chapter describes some research works done in the literature to parallelize a B&B algorithm using threads. It finally presents the multithreaded B&B approach proposed for solving the flow-shop problem on a multicore system based on POSIX threads, and reports the obtained experimental results. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Large-Scale P2P-Inspired Problem-Solving: A Formal and Experimental Study

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch04
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter focuses on parallel/distributed branch-and-bound (B&B) techniques for solving very large instances requiring the use of a huge amount of computational resources. First, it provides some background concerning the parallelization of the B&B algorithm. Next, the chapter describes the peer-to-peer (P2P) approach and details the main distributed challenges. It further focuses on an analysis of P2P approach from scalability and congestion perspectives. The chapter also provides experimental results on an instance of the flow-shop problem (Ta057), comparing the approach with the state-of-the-art one. A detailed analysis of the intrinsic performances of the approach is provided through the study of the different execution phases. Finally, the chapter lists some open questions raised by the P2P approach. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Data Distribution Management

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch05
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      Data distribution management (DDM) deals with two basic problems: how to distribute data generated at the application layer among underlying nodes in the distributed system and how to retrieve data whenever it is necessary. DDM is addressed in two different network environments: peer-to-peer (P2P) overlay networks and cluster-based network environments. Although DDM provides the same data indexing service in both distributed network environments, it serves upper-layer applications for very different purposes, and it is supported by underlying networks with distinctive infrastructures. This chapter introduces some basic concepts that are widely used in the high-level architecture (HLA)/DDM based distributed simulations community. In order to understand these concepts more thoroughly, their different representations as used by two major DDM systems: the region-based system and the grid-based system are incorporated. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Middleware Support For Context Handling and Integration in Ubiquitous COMPUTING

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch06
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      Ubiquitous computing (UC) encompasses sensor-instrumented environments, which are often endowed with wireless network interfaces, in which devices, software agents, and services are integrated in a seamless and transparent way and cooperate to meet high-level goals of human users. This chapter introduces some background concepts on UC. It presents the state-of-the-art for the field of middleware for UC. The requirement of the integration of UC middleware is addressed and an existent solution is described - the Open Context Platform Integration (OpenCOPI) platform. OpenCOPI provides an automatic service composition, a standardized context and communication model, and a fault tolerance mechanism that handles service failures. Another important feature of OpenCOPI is the use of semantic workflows. Finally, the chapter presents a new integration platform for UC. It talks about OpenCOPI as an example of a software system developed to integrate different context provision middleware, with the goal of facilitating the development of context-aware ubiquitous applications. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Challenges In The Use of Wireless Sensor Networks for Monitoring The Health of Civil Structures

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch07
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter provides an overview of the challenges faced in the design of new techniques for enabling new decentralized solutions of large-scale wireless sensor networks (WSNs) in the structural health monitoring (SHM) domain. It introduces the definition of SHM along with the concept of WSNs. The chapter discusses the concepts of SHM and WSNs apart from one another. The chapter explains existent solutions employing WSNs in the context of SHM. It focuses on SHM techniques based on the use of accelerometers; but in further investigations, this classification can be expanded to works that use other kinds of sensing devices, for example, strain gauges, following the same logic of higher degrees of decentralization and in-network processing. The concept of generations of sensor networks for SHM was used for such classification. Each generation is presented by describing respective examples of works found in the current literature. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Mobility Effects In Wireless Mobile Networks

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch08
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter discusses the effect of mobility on link and topology of wireless networks. A common approach for analysis of the effect of mobility on the link lifetime (LL) and residual link lifetime (RLL) of wireless networks is geometrical modeling. Using the geometric model, probability density function of LL or RLL can be obtained for simple synthetic mobility models. For analysis of the effect of mobility on topology of wireless networks, a phase transition phenomenon was observed for different scenarios for two well-known topology control protocols, homogeneous topology control and k-Neigh topology control. The authors observed that disconnection occurs suddenly in the networks rather than gradually. Another observation was that the degree of the effective topology was almost fixed around 4.5 at the disconnection time. Using these facts, one can evaluate the lifetime of the topologies which could be directly used for optimizing the hello interval in topology control protocols. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Analytical Model of Time-Critical Wireless Sensor Network: Theory and Evaluation

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch09
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter establishes a probabilistic fundamental quantitative notion for time-critical applications on real-time information transfer in multihop wireless networks. Those wireless sensor networks that are capable of providing bounded delay guarantees on packet delivery are referred to as real-time wireless sensor networks (RT-WSN). The preliminaries and assumption of the problem are clarified. The chapter derives a model for evaluating the reliable real-time (RRT) degree and presents the model validation for RT-WSN. The model can be extended to describe various aspects in the design of sensor networks, such as data aggregation or backpressure traffic mechanisms. Finally, the chapter deals with the cluster-based network architectures as well as the case where the network topology varies. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Multicast Transport Protocols for Large-Scale Distributed Collaborative Environments

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch10
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter provides an overview of multicast transport protocols. It also provides an introduction to multicasting and the features that can be added to satisfy each application's requirements. The chapter describes multicast protocols by classifying them according to several features. As the number of applications using multicasting has grown, the number of multicast transport protocols has increased as well. The latter can be classified into three categories, depending on the type of applications: general-purpose protocols such as reliable broadcast protocol (RBP), multicast transport protocol (MTP), reliable multicast protocol (RMP), and Xpress transport protocol (XTP); multicast interactive applications such as multicast transport protocol-2 (MTP-2), real-time transport protocol (RTP), scalable reliable multicast (SRM), and reliable adaptive multicast protocol (RAMP); and data distribution services such as tree-based multicast transport protocol (TMTP), reliable multicast transport protocol (RMTP), multicast file transfer protocol (MFTP), and tree-based reliable multicast protocol (TRAM). View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Nature-Inspired Computing for Autonomic Wireless Sensor Networks

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch11
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter presents how nature-inspired algorithms can affect the underlying design principles, enabling technologies, implementation, and management issues of autonomic wireless sensor networks (WSNs). It explains how WSN can benefit from autonomic mechanisms. The chapter provides a brief introduction to nature-inspired computing. It reviews several current nature-inspired techniques such as genetic algorithms (GA) and genetic programming (GP) to construct autonomic WSN. The chapter also presents the state-of-the-art techniques in those areas. It introduces several works such as bio-networking architecture, and biologically-inspired architecture for sensor NETworks (BiSNET) in which multiple biological principles are used to build fundamental frameworks for implementing autonomic WSNs. Self-organized mechanisms alleviate the dependence of sensors on a central controller and therefore help sensors to consume energy more efficiently in reaching the primary goal of WSNs to prolong their lifetime. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Smart RPC-Based Computing in Grids and on Clouds

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch12
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      The remote procedure call (RPC) paradigm is widely used in distributed computing. It provides a straightforward procedure for executing parts of an application on a remote computer. The aim of the GridRPC model is to provide a standardized, portable, and simple programming interface for a RPC. SmartGridRPC is an enhancement of the traditional GridRPC model that allows a group of tasks to be mapped collectively onto a fully connected network. Cloud computing allows users to have easy access to large distributed resources from data centers anywhere in the world. This chapter describes the three major categories of cloud computing. These are Infrastructure as a Service (IaaS), Platform as a Service (PaaS), which is the category the smart cloud solve (SCS) platform falls under, and Software as a Service (SaaS). View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Profit-Maximizing Resource Allocation for Multitier Cloud Computing Systems Under Service Level Agreements

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch13
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter reviews the important approaches and techniques for addressing power and performance management problems in a datacenter. Next, it discusses a problem related to maximization of the profit in a hosting datacenter or cloud computing system and proposes an efficient solution. A review of some of the approaches and techniques addressing power management and performance management is presented. A system model for the profit maximization problem is explained. The chapter describes the problem formulation and the proposed solution. A model based on the multitier applications is presented and the guarantee-based service level agreement (SLA) is used to model the profit in the system. Two models of the SLA are considered: (i) the Gold SLA class and (ii) Bronze SLA class. The chapter concludes with simulation results. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Market-Oriented Cloud Computing and The Cloudbus Toolkit

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch14
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter introduces the fundamental concepts of market-oriented Cloud computing systems and presents a reference model. The model, together with the state-of-the-art technologies presented in the chapter, contribute significantly towards the mainstream adoption of Cloud computing technology. However, any technology brings with it new challenges and breakthroughs. The chapter focuses on the major challenges faced by the industry when adopting Cloud computing as a mainstream technology as part of the distributed computing paradigm. It presents a utility-oriented Cloud vision that is a generic model for realizing market-oriented Cloud computing vision. Cloudbus realized this by developing various tools and platforms that can be used individually or together as an integrated solution. The author's demonstrate through experiments that their toolkit can provide applications based on deadline, optimize cost and time of applications, and manage real-world problems through an integrated solution. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      A Cloud Broker Architecture for Multicloud Environments

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch15
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter presents a cloud broker architecture oriented to manage the deployment of multitier and elastic services among available cloud providers. The chapter describes the current state of cloud brokers and their features. It analyzes the implementation of cloud brokering mechanisms representing several challenges. The chapter depicts the cloud broker architecture for multicloud environments and its components, specially focusing on the service description. It explains the scheduling policies used by the broker. The chapter discusses the obtained scheduling results in a single cloud scenario, presenting a use case in which the broker reduces the overall cost: deployment of a grid in a single cloud. It also discusses the obtained scheduling results in a multi-cloud scenario, showing the benefit of using the broker in two use cases: a web server and a virtual classroom. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Energy-Efficient Resource Utilization in Cloud Computing

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch16
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      In Cloud computing, energy consumption and resource utilization are strongly coupled. Specifically, resources with a low utilization rate still consume an unacceptable amount of energy compared to the energy consumption of a fully utilized or sufficiently loaded Cloud computing. To increase resource utilization, task consolidation is an effective technique, greatly enabled by virtualization technologies, which facilitate the concurrent execution of several tasks and, in turn, reduce energy consumption. The authors use two energy-conscious heuristics for task consolidation: MaxUtil, which aims to maximize resource utilization, and Energy-Conscious Task Consolidation (ECTC) which explicitly takes into account both active and idle energy consumption. ECTC computes the energy consumption based on an objective function derived from findings reported in the literature. The chapter deals with the Cloud computing, energy models, and task consolidation algorithms. It describes the complementarity approach and the related mathematical model. The chapter finally summarizes the simulation results and discussions. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Semantics-Based Resource Discovery in Large-Scale Grids

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch17
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      Semantic virtual organizations (VOs) help to discriminatively distribute resource information and queries to related nodes, thus reducing the search space and improving scalability. This chapter proposes a semantics-based searching mechanism, OntoSum, in order to further improve the efficiency of searching the virtual organizations. A section provides a general overview of related work in resource discovery in grids, and another describes the approach to the construction of semantic VOs. This is followed by a presentation on OntoSum, a framework for semantic resource discovery in virtual organizations. The implementation of a prototype system is discussed, and finally, the authors' conclusions, discussion of limitations, and suggestions for future work are presented. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Game-Based Models of Grid User's Decisions in Security-Aware Scheduling

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch18
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter summarizes the recent work of the authors on the game-theoretic models of the grid users behavior in security aware scheduling. The scheduling process may be realized in two alternate scenarios, namely risky and secure modes. The chapter focuses on the fundamental type of the scheduling problem in computational grids, namely independent batch scheduling, where it is assumed that tasks are clustered and grouped into batches and there is no inter-relation among them. It considers Makespan and Flowtime as the main scheduling criteria expressed in terms of the completion times of the machines by using the ETC matrix model. The chapter focuses on the non-zero sum games, where the strategies of the players are not opposite, that is, the sum or the values of all players cost functions is not zero. Player's Genetic Algorithm (PGA) and Player's Minimum Completion Time (PMCT) are presented as case studies. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Addressing Open Issues on Performance Evaluation in Cloud Computing

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch19
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter discusses the roles of benchmarking and monitoring of service performance in Cloud computing. The chapter talks about the effects of cyber attack on Cloud service provisioning. It presents a survey on the proposed solutions and technologies to face such issues. The chapter describes benchmark results and products from different points of view. It offers a brief critical analysis of proposed solutions, and outlines the open research problems. Evaluation and monitoring of provided services allows for an accurate assessment of system performance, workload distribution, overloads, bottleneck, underused resources, and malicious resource exhaustion, which have special effects in the Cloud because of the pay-per-use business model. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Broker-Mediated Cloud-Aggregation Mechanism Using Markovian Queues For Scheduling Bag-of-Tasks (BoT) Applications

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch20
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter discusses a Broker-mediated Cloud Aggregation mechanism for Bag-of-Tasks (BoT) applications wherein users can submit their applications along with the task execution requirements to the Broker and the Broker distributes the tasks to various Cloud Service Providers (CSPs) based on these requirements. The chapter briefly describes the relevant literature on Cloud brokering mechanisms, and models for deploying tasks on multiple Clouds. The problem setting and notations are explained and the proposed Cloud aggregation mechanism is detailed. The chapter focuses on the Monkovian queue-based optimization problem and derives optimal load fractions for task distribution in order to minimize task execution time. It further explains a heuristic algorithm to take into account the user's budget constraints. The performance evaluation and results are explained. The chapter deals with some important points about the proposed model, and finally provides some future insights. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      On The Design of A Budget-Conscious Adaptive Scheduler for Handling Large-Scale Many-Task Workflow Applications in Clouds

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch21
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter presents a budget-conscious adaptive scheduler, which minimizes many-task workflow (MTW) execution time within a certain budget. The proposed MTW scheduling approach can be used by both end users and Cloud Service Providers (CSPs). End users can use the approach to orchestrate Cloud services, whereas CSPs can outsource computing resources to meet customers' service-level requirements. The chapter addresses the problem of scheduling MTW applications in a Cloud. It provides the system model and schedule problem definition and also presents both static and dynamic strategies for MTW scheduling. The chapter evaluates the performance of algorithms and discusses how efficiently they can adapt to a Cloud system dynamically. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Virtualized Environment Issues in The Context of A Scientific Private Cloud

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch22
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter presents an approach using the virtualization advancements, and the knowledge obtained by tests results to verify the use of hybrid virtual environments, that is, composed of two types of hypervisors/virtual machine management (VMM), to support scientific applications. These hybrid models are created based on characteristics of the running applications, combining the use of different types of hypervisors, at the same time, using the best features of each one, to improve the execution performance. The evaluation of the use of these hybrid models and their analysis were motivated by a real demand, to provide access to a Cloud Appliance with a graphical interface (GUI) for distributed image-processing applications. The basic idea is to offer users remote access to virtualized platforms dedicated to their needs and hosted in a Cloud environment. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      In-Advance Bandwidth Scheduling In E-Science Networks

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch23
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      In reality, on-demand scheduling is a special case of in-advance bandwidth scheduling; the future time for each scheduling request is separated from the time at which the request is made by a time interval of zero. Hence, this chapter considers only the in-advance scheduling mode. The development is naturally adapted to the on-demand mode. Much of the research on bandwidth scheduling has focused on reserving a single path for a specified bandwidth request. For on-demand scheduling, this is typically supported by Multiple Protocol Label Switching (MPLS) at layer three and by Generalized MPLS (GMPLS) at layers one and two. The chapter describes temporal network model. In-advance scheduling problems are defined and corresponding algorithms are proposed and evaluated. A multiple path scheduling problem Earliest Finish Time File Transfer Problem (EFTFTP) is proposed and solved by both optimal solutions and heuristics. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Routing and Wavelength Assignment in Optical Networks

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch24
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      The algorithms presented in this chapter assume that either wavelength conversion is available at all switches or not available at any switch. The algorithms assume that the wavelength sharing constraint needs to be satisfied. The Extended Bellman-Ford (EBF) algorithm is extended in the chapter in order to incorporate the wavelength sharing and wavelength continuity constraints. The authors propose modified versions of the algorithms in the Enlightened system for continuous time model. These algorithms are called Modified Switch Path First (MSPF) and Modified Switch Window First (MSWF). The chapter presents a new network model that can emulate the existing full-conversion algorithms when only a subset of nodes have a wavelength converter. The authors demonstrate the utility of this approach using the EBF and k-Alternative Path (k-Path) algorithms. The chapter evaluates the algorithms on three performances metrics: blocking probability, average start time and scheduling overhead. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Computational Graph Analytics for Massive Streaming Data

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch25
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      In this chapter the author presents a new, extensible and flexible data structure for massive graphs called STINGER (Spatio-Temporal Interaction Networks and Graphs (STING) Extensible Representation). Two studies are discussed: the first study, computing a widely used network analysis metric called clustering coefficients, and the second study, the approach for tracking connected components given a stream of edge insertions and removals. The chapter describes the massively multithreaded Cray XMT as well as the more common platform built on Intel's Nehalem architecture. It presents experimental results on the Cray XMT, with a comparison to the Intel Nehalem platform for the clustering coefficient problem. Regarding clustering coefficients, a similar rate of 200,000 updates per second can be maintained on the Cray XMT. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Knowledge Management For Fault-Tolerant Water Distribution

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.ch26
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      This chapter presents and validates environmental decision support systems (EDSSs) where ontologies are utilized to classify failure events and determine appropriate countermeasures to mitigate their effects. It provides an overview of previous related work in EDSSs, agent-based modeling and simulation techniques, knowledge representation capabilities of ontology, and simulation techniques as tools for validation of models for cyber-physical system (CPS). The chapter discusses the major functional components of a generic CPS, the major components of the derivative water distribution network (WDN), and the agent-based model for WDN operation. It also provides a detailed introduction to the class composition of the ontology-based framework. The chapter illustrates the automated failure classification and mitigation. It outlines the object properties for behavior reasoning and the data properties for value reasoning, respectively. Finally, the chapter demonstrates the benefits of the ontology-based framework through a test case on an integrated CPS simulator for WDN. View full abstract»

    • Full text access may be available. Click article title to sign in or learn about subscription options.

      Index

      Sarbazi-Azad, H. ; Zomaya, A.
      Large Scale Network-Centric Distributed Systems

      DOI: 10.1002/9781118640708.index
      Copyright Year: 2014

      Wiley-IEEE Press eBook Chapters

      No abstract. View full abstract»