Cérin / Cerin / Li Advances in Grid and Pervasive Computing

Second International Conference, GPC 2007, Paris, France, May 2-4, 2007, Proceedings
2007
ISBN: 978-3-540-72360-8
Verlag: Springer Berlin Heidelberg
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 776 Seiten

Reihe: Theoretical Computer Science and General Issues

This book constitutes the refereed proceedings of the Second International Conference on Grid and Pervasive Computing, GPC 2007, held in Paris, France in May 2007.

The 56 revised full papers and 12 revised short papers were carefully selected from 217 submissions during two rounds of reviewing and improvement. The papers address all aspects of grid and pervasive computing and focus on topics such as cluster computing, high performance computing, grid computing, semantic Web and semantic grid, service-oriented computing, peer-to-peer computing, pervasive computing, mobile computing, network storage, as well as grid and pervasive related applications.

Written for: Researchers and professionals

Keywords: QoS, Web services, ad-hoc networks, biometrics, cluster computing, data grid, distributed algorithms, distributed computing, distributed processing, distributed systems, fault tolerance, fuzzy logic, grid computing, grid scheduling, high performance computing, interoperability, load balancing, middleware, mobile systems, multicast protocols, network computing, parallel processing, peer-to-peer systems, performance modeling, pervasive computing, reliability, routing, semantic Web, sensor networks, ubiquitous computing, user interfaces, wireless networking.

Autoren/Hrsg.

Cérin, Christophe

Cerin, Christophe

Li, Kuan-Ching

Inhaltsverzeichnis

1;Preface;6
2;Organization;8
3;Table of Contents;12
4;A Grid Resource Broker with Network Bandwidth-Aware Job Scheduling for Computational Grids;18
4.1;Introduction;18
4.2;Related Work;19
4.3;Design and Implementation of Resource Broker;20
4.4;Design of Network Bandwidth-Award Job Scheduling;22
4.4.1;Mechanism of Performance Evaluation;22
4.4.2;The Algorithm;23
4.5;Experimental Environment and Results;25
4.6;Conclusions;28
4.7;References;28
5;Design of PeerSum: A Summary Service for P2P Applications;30
5.1;Introduction;30
5.2;PeerSum Summary Model;31
5.2.1;Model Architecture;32
5.2.2;Summarization Process: Scalability Issues;33
5.2.3;Summary Representation;33
5.3;Summary Management in PeerSum;34
5.3.1;Summary Construction;35
5.3.2;Summary Maintenance;36
5.3.3;Peer Dynamicity;37
5.4;Query Processing;38
5.4.1;Query Extension;38
5.4.2;Query Evaluation;39
5.5;Performance Evaluation;39
5.5.1;Cost Model;40
5.5.2;Discussion;41
5.6;Conclusion;42
6;A High-Performance Virtual Storage System for Taiwan UniGrid;44
6.1;Introduction;44
6.2;System Framework and Deployment;46
6.3;Main Features;48
6.3.1;Multi-source Data Transfer;48
6.3.2;Date Sharing;49
6.3.3;Single Sign-On;50
6.3.4;The Data Management Client;52
6.4;Operation Scenario of Taiwan UniGrid;53
6.5;Conclusion;54
6.6;References;55
7;Interoperable Grid PKIs Among Untrusted Domains: An Architectural Proposal;56
7.1;Introduction;56
7.2;State of the Art;58
7.2.1;Grid Validation;58
7.2.2;Evaluation of Grid PKIs;59
7.3;The Problem of Grid Security Interoperability;60
7.4;The Architectural Model of an Interoperability System;61
7.5;POIS: Policy and OCSP Based Interoperability System;63
7.5.1;Extended Path Validation: POIS and the End-Entity;64
7.5.2;Extended Path Validation: POIS and the Grid Services Container;65
7.6;Conclusions and Future Work;66
7.7;References;67
8;TCMM: Hybrid Overlay Strategy for P2P Live Streaming Services;69
8.1;Introduction;69
8.2;Related Works;70
8.3;Design of TCMM;71
8.3.1;Overview of TCMM;71
8.3.2;Tree Management;71
8.3.3;Mesh Management;73
8.4;Performance Evaluation;75
8.4.1;Simulation Setup;75
8.4.2;Control Overhead;76
8.4.3;Starting Delay;77
8.4.4;Dynamic Resistance;77
8.5;Conclusions;78
8.6;References;79
9;Fault Management in P2P-MPI;81
9.1;Introduction;81
9.2;P2P-MPI Overview;82
9.3;Replication and Failure Probability;84
9.4;Fault Detection: Background;85
9.5;Fault Detection in P2P-MPI;87
9.5.1;Assumptions and Requirements;88
9.5.2;Design Issues;89
9.5.3;P2P-MPI Implementation;89
9.6;Experiments;92
9.7;Conclusion;93
10;Heterogeneous Wireless Sensor Network Deployment and Topology Control Based on Irregular Sensor Model;95
10.1;Introduction;95
10.2;Related Work;96
10.3;Preliminaries;97
10.3.1;Irregular Sensor Model;97
10.3.2;Some Definitions of Heterogeneous Wireless Sensor Network;98
10.4;Heterogeneous Sensor Deployment;99
10.4.1;Initialization Step;100
10.4.2;Neighbor-Info Collection Step;100
10.4.3;Candidates Generation Step;101
10.4.4;Scoring Step;102
10.4.5;Sensor Addition Step;102
10.5;Experiments;103
10.6;Conclusions;104
10.7;References;105
11;Multiple Cluster Merging and Multihop Transmission in Wireless Sensor Networks;106
11.1;Introduction;106
11.2;Uniform Cluster-Head Distribution;108
11.2.1;Cluster Merging;109
11.3;2-Level LEACHM;111
11.3.1;Master Cluster-Head Determination;112
11.4;Simulation Results and Analysis;113
11.4.1;Sensitivity Analysis of LEACHM;114
11.5;Conclusion;115
12;CFR: A Peer-to-Peer Collaborative File Repository System;117
12.1;Introduction;117
12.2;Related Work;118
12.3;System Overview;119
12.4;The Overlay Management of CFR;120
12.4.1;The Base Overlay;120
12.4.2;The Region Overlay;120
12.5;The File Management of CFR;122
12.5.1;Insert Files and Create Duplicates in CFR;122
12.5.2;Retrieve and Remove Files in CFR;123
12.5.3;Dealing with Storage Node Dynamics;123
12.6;Simulation Results;124
12.6.1;Expected Number of Hops to Collect All Links;124
12.6.2;Evaluation of File Management of CFR;125
12.7;Experimental Results;125
12.8;Conclusions and Future Work;127
12.9;References;127
13;Optimal Deployment of Mobile Sensor Networks and Its Maintenance Strategy;129
13.1;Introduction;129
13.2;Technical Preliminaries;130
13.2.1;Fuzzy Logic Systems;130
13.2.2;Coverage;131
13.3;Proposed Deployment Approach: EFOA;132
13.3.1;Assumptions and Model;132
13.3.2;Energy-Efficient Fuzzy Optimization Algorithm;132
13.4;Proposed Network Maintenance Strategy;136
13.5;Performance Evaluations;137
13.6;Conclusions and Future Work;139
13.7;References;139
14;Server Placement in the Presence of Competition;141
14.1;Introduction;141
14.2;Problem Formulation;143
14.3;Finding Extra Server Locations;144
14.4;NP-Completeness;148
14.5;Experiment Results;148
14.5.1;Experiment Setting;149
14.5.2;Effect of;150
14.5.3;Effect of the Number of Original Servers;150
14.5.4;Effect of k;150
14.6;Conclusion;151
15;A Scalable Mechanism for Semantic Service Discovery in Multi-ontology Environment;153
15.1;Introduction;153
15.2;Related Works;154
15.3;Architecture of SSD_OC;154
15.4;Semantic Service Matching in SSD_OC;155
15.4.1;Related Definitions;155
15.4.2;Semantic Service Matching in SSD_OC;158
15.4.3;Put Them Together-A Service Discovery Example;159
15.5;Experiments;160
15.6;Conclusions and Future Works;161
15.7;References;162
16;A Collaborative-Aware Task Balancing Delivery Model for Clusters;163
16.1;Introduction;163
16.2;Related Work;164
16.3;CAMT: Reinterpreting the Key Awareness Concepts;164
16.4;Load Balancing Algorithm in CAMT;166
16.4.1;State Measurement Rule;166
16.4.2;Information Exchange Rule;167
16.4.3;Initiation Rule;167
16.4.4;Load Balancing Operation;167
16.5;The CAMT Architecture;168
16.5.1;The Load Agent;168
16.5.2;The Global State Agent;169
16.5.3;The Initiation Agent;169
16.5.4;The Load Balancer Agent;170
16.6;Experimental Results;170
16.7;Conclusions;173
16.8;References;174
17;An Improved Model for Predicting HPL Performance;175
17.1;Introduction;175
17.2;HPL Algorithm and Performance Score Model;176
17.2.1;HPL Algorithm;176
17.2.2;Performance Score Model;177
17.3;Comparative Analysis of Different Models on Various Clusters;179
17.3.1;NCHC Formosa PC Cluster;180
17.3.2;NCHC Triton Cluster;181
17.3.3;Dawning 4000A;182
17.4;Prediction of $R_max$ on SIRAYA;183
17.5;Conclusion;184
17.6;References;184
18;An Ad Hoc Approach to Achieve Collaborative Computing with Pervasive Devices;186
18.1;Introduction;186
18.2;Adaptive Software Framework: FRAME;188
18.3;Reassembly;190
18.4;Performance Evaluation;192
18.5;Related Work;194
18.6;Conclusion and Future Work;195
19;Optimizing Server Placement for QoS Requirements in Hierarchical Grid Environments;198
19.1;Introduction;198
19.2;The System Model;199
19.3;The Minimum Server Placement Problem;200
19.4;The Optimal Service Quality Problem;205
19.5;Experimental Results;206
19.6;Conclusions;208
20;AHSEN – Autonomic Healing-Based Self Management Engine for Network Management in Hybrid Networks;210
20.1;Introduction;210
20.2;Related Work;212
20.3;Proposed Architecture;213
20.3.1;Software Architecture;213
20.4;Implementation Details;217
20.5;Concluding Remarks;218
20.6;References;219
21;Development of a GT4-Based Resource Broker Service: An Application to On-Demand Weather and Marine Forecasting;221
21.1;Introduction;221
21.2;The Resource Broker Architecture and Design;223
21.3;The Native Latent Semantic Indexing Based Matchmaking Algorithm;225
21.4;The Condor ClassAd Based Matchmaking Algorithm;227
21.5;An Application to on Demand Weather and Marine Forecasting;227
21.6;Conclusions and Future Works;232
21.7;References;232
22;Small-World Network Inspired Trustworthy Web Service Evaluation and Management Model;235
22.1;Introduction;235
22.2;Small-World Network and WS Trustworthiness;236
22.2.1;Trustworthiness in the Human Small-World Network;236
22.2.2;WS Trustworthiness;236
22.3;WS Trustworthiness Evaluating and Managing Model;237
22.3.1;WS Trustworthy Management Model;237
22.3.2;APAEAS and AWSORT;238
22.4;Web Service Federation Organization Protocols, Data Structures and Instructions;239
22.4.1;AWSROT Protocol;239
22.4.2;APAEAS Protocol;239
22.4.3;Instruction System;241
22.5;WS Evaluating and Managing Algorithm;241
22.5.1;WS Assigning Algorithm;241
22.5.2;WS Loading-Balance Algorithm;242
22.6;Simulations of Trustworthy WS Assigning;243
22.7;Conclusions;244
22.8;References;245
23;Towards Feasible and Effective Load Sharing in a Heterogeneous Computational Grid;246
23.1;Introduction;246
23.2;Related Work;247
23.3;Computational Grid Model and Experimental Setting;248
23.4;Site Selection Policies for Load Sharing in a Heterogeneous Grid;250
23.5;Feasible Load Sharing in a Computational Grid;253
23.6;Multi-site Parallel Execution in a Heterogeneous Grid;254
23.7;Conclusion;255
23.8;References;256
24;Meeting QoS Requirements of Mobile Computing by Dual-Level Congestion Control;258
24.1;Introduction;258
24.2;Research Background;259
24.3;UBC-CAC System Model;260
24.3.1;Utility Function;260
24.3.2;System Model;260
24.4;Design of UBC Module;261
24.4.1;Congestion Detection;261
24.4.2;User Traffic Shaper;262
24.4.3;User Notification;262
24.5;User Behavior Modes and Performance Evaluation;263
24.5.1;User Behavior Modes;263
24.5.2;Simulation;265
24.6;Conclusions;267
24.7;References;268
25;A Transaction Model for Context-Aware Applications;269
25.1;Introduction;269
25.2;Related Work;270
25.3;Motivation and Further Analyses;271
25.3.1;Cause for Anomalies in Context-Aware Service Providing;271
25.3.2;Necessity and Benefits of Applying Transaction Models;271
25.3.3;Requirements of Transaction Model for Context-Awareness;273
25.4;A Transaction Model: TMfm;273
25.4.1;Formalizing of Context-Aware Applications;273
25.4.2;Scopes;274
25.4.3;TMfm Model;276
25.5;An Implementation of TMfm;276
25.6;Discussion;278
25.7;Conclusion;279
25.8;References;279
26;A Grid-Based Remote Experiment Environment in Civil Engineering;280
26.1;Introduction;280
26.2;Related Works;281
26.3;KOCEDgrid;282
26.4;Collaborative Research Environment;284
26.4.1;Remote Experiment;284
26.4.2;Collaborative Environment;285
26.5;Supporting Remote Experiment;285
26.5.1;Hybrid Test Model;286
26.5.2;Building a Prototype for Hybrid Test Model;286
26.6;Conclusions;289
27;Mobile Ad Hoc Grid Using Trace Based Mobility Model;291
27.1;Introduction;291
27.2;Related Work;292
27.3;Proposed Architecture for Mobile Ad Hoc Grid;293
27.3.1;Grid Formation;294
27.4;Mobile Ad Hoc Grid Evaluation;298
27.5;Conclusion and Future Work;301
27.6;References;301
28;Self Managing Middleware for Dynamic Grids;303
28.1;Introduction;303
28.2;Self Managed Grid Middleware;304
28.2.1;Software Manager;306
28.2.2;Security Manager;307
28.2.3;Software Repository;307
28.3;Resource Manager;308
28.3.1;Resource Mapping;308
28.3.2;Architecture;309
28.3.3;Federation of Registry;311
28.4;Use Case;312
28.5;Conclusion;313
28.6;References;313
29;Adaptive Workflow Scheduling Strategy in Service-Based Grids;315
29.1;Introduction;315
29.2;Problem Statement;316
29.2.1;Task Graph;316
29.2.2;Service Graph;317
29.2.3;Performance Criteria;317
29.3;Adaptive Scheduling Using Dynamic Maximum Flow Algorithm;320
29.4;Experiment;322
29.4.1;Performance Evaluation According to the Number of Nodes for Services;322
29.4.2;Performance Evaluation According to the Number of Tasks;323
29.4.3;Performance Evaluation in Real Grid Application;323
29.5;Related Works;324
29.6;Conclusion;325
29.7;References;325
30;Scalable Thread Visualization for Debugging Data Races in OpenMP Programs;327
30.1;Introduction;327
30.2;Background;328
30.2.1;OpenMP Program;328
30.2.2;Race Detection Tools;330
30.3;Scalable Thread Visualization;331
30.3.1;Thread Visualization;331
30.3.2;Scalable Visualization;332
30.4;Experimentation;334
30.4.1;Visualization Engines;334
30.4.2;Visualization Cases;335
30.5;Conclusion;337
31;MPIRace-Check: Detection of Message Races in MPI Programs;339
31.1;Introduction;339
31.2;Background;340
31.2.1;Message Races;341
31.2.2;Related Work;341
31.3;Race Detection;344
31.3.1;Concurrency Check;344
31.3.2;MPI Profiling Interface;346
31.4;Experimentation;347
31.5;Conclusion;349
32;The Modified Grid Location Service for Mobile Ad-Hoc Networks;351
32.1;Introduction;351
32.2;Overview of MGLS Scheme;353
32.2.1;Grid Hierarchy;354
32.2.2;Location Servers;354
32.2.3;Design Tradeoffs;356
32.3;Comparisons Based on a Theoretical Model;357
32.3.1;Metrics;357
32.3.2;Model Assumptions;357
32.3.3;MGLS;358
32.3.4;GLS;359
32.3.5;Summary of Theoretical Analyses;360
32.4;Performance Evaluation Using Simulation;361
32.5;Conclusions;363
33;Authentication and Access Control Using Trust Collaboration in Pervasive Grid Environments;365
33.1;Introduction;365
33.2;Pervasive Grid Scenario;366
33.3;The Chameleon Architecture;367
33.3.1;The Architecture Description;368
33.4;How Foreign User Accesses Unknown Site?;369
33.4.1;Requirements;370
33.4.2;Chameleon Behavior;373
33.5;Implementation and Discussion;376
33.6;Conclusion;377
34;Architecture-Based Autonomic Deployment of J2EE Systems in Grids;379
34.1;Introduction;379
34.2;Context and Main Assumptions;380
34.2.1;J2EE System Configuration and Deployment;380
34.2.2;Deployment in a J2EE Cluster;381
34.2.3;From J2EE Clusters Management to Virtual Clusters Management;381
34.3;Virtual Cluster Deployment System;382
34.3.1;Deployment Specification;383
34.3.2;Deployment Process;383
34.3.3;Automatic Recovery from Failures;385
34.4;Implementation Status and Evaluation;386
34.4.1;Implementation Status;386
34.4.2;Evaluation;387
34.5;Related Work;388
34.6;Conclusion;389
35;Dynamic Workload Balancing for Collaboration Strategy in Hybrid P2P System;391
35.1;Introduction;391
35.2;Related Works;392
35.3;Dynamic Workload Management;393
35.3.1;Importance of SP Workload;393
35.3.2;Workload Value Evaluation;394
35.3.3;Workload Status Classification;395
35.3.4;Collaboration Policy;395
35.4;Experimental Evaluations;397
35.5;Conclusion;400
35.6;References;401
36;Performance-Based Workload Distribution on GridEnvironments;402
36.1;Introduction;402
36.2;Background Review;403
36.2.1;Dynamic Loop Scheduling Schemes;403
36.2.2;Association Rule Mining;404
36.3;Approach: Performance-Based Workload Distribution (PWD);404
36.3.1;The System Model;405
36.3.2;Performance Ratio;405
36.3.3;Determination of Static-Workload Ratio (SWR);406
36.3.4;Algorithm;407
36.4;Experimental Results;408
36.4.1;Application 1: Matrix Multiplication;409
36.4.2;Application 2: Association Rule Mining;409
36.4.3;Application 3: Mandelbrot Set Computation;410
36.5;Conclusions;411
36.6;References;411
37;A Visual Framework for Deploying and Managing Context-Aware Services;414
37.1;Introduction;414
37.2;Background;415
37.2.1;Symbolic Location Model;415
37.2.2;Compound Document-Based Management Interface;416
37.2.3;Basic Approach;416
37.2.4;Remarks;417
37.3;M-Space: Location Model for Smart Spaces;417
37.3.1;Containment Relationship Model;418
37.3.2;Agent;418
37.4;Compound Document Framework for Managing Pervasive Computing;420
37.4.1;Visual Component;420
37.4.2;Component Runtime System;420
37.5;Binding Between Visual Components and Virtual Counterparts;422
37.5.1;Updating the Structure and Attributes of Visual Components;422
37.5.2;Updating the Structure and Attributes of Agents;422
37.6;Early Experience;424
37.6.1;Management System for Context-Aware Services;425
37.7;Related Work;426
37.7.1;Location Models;426
37.7.2;Management Systems for Pervasive Computing;426
37.8;Conclusion;427
38;Towards a Peer-To-Peer Platform for High Performance Computing;429
38.1;Introduction;429
38.2;$XtremWeb-CH$ Ingredients;430
38.2.1;The Coordinator;431
38.2.2;The Workers;432
38.2.3;The Warehouses;433
38.2.4;The Brokers;433
38.3;$XWCH$ Characteristics;434
38.3.1;Support of Communicating Tasks;434
38.3.2;Direct Communication Between Workers;434
38.3.3;Granularity and Scheduling;435
38.4;Experiments;436
38.4.1;The Application;437
38.4.2;Evaluation of the Scheduling Algorithm;438
38.5;Conclusion;439
38.6;References;439
39;Assessing Contention Effects on MPI_Alltoall Communications;441
39.1;Introduction;441
39.2;Related Works;442
39.3;Network Models Definition;444
39.4;Problem Definition;444
39.4.1;Notation and Lower Bounds;445
39.5;Contention Signature Approach;445
39.5.1;Non-linear Aspects;446
39.6;Validation;447
39.6.1;Fast Ethernet;447
39.6.2;Gigabit Ethernet;448
39.6.3;Myrinet;448
39.7;Applications to Grid-Aware Communications;450
39.7.1;Performance Prediction in a Grid Environment;450
39.8;Conclusions and Future Works;451
40;An Energy-Efficient Clustering Algorithm for Large-Scale Wireless Sensor Networks;453
40.1;Introduction;453
40.2;Related Works;455
40.3;SNOWCLUSTER Algorithm;456
40.4;Performance Evaluation;457
40.4.1;Simulation Environments;457
40.4.2;Energy Consumption;458
40.4.3;The Amount of Data;461
40.4.4;Network Lifetime;462
40.5;Conclusion;463
41;An Algorithm Testbed for the Biometrics Grid;464
41.1;Introduction;464
41.2;Related Work;466
41.3;Design Issues of System;467
41.3.1;Concepts;467
41.3.2;A Framework of the Biometrics Grid;467
41.3.3;The BMG Workflow for the Algorithm Testbed;469
41.4;A Testbed for Biometric Algorithms;469
41.4.1;Single Biometric Test;470
41.4.2;Multimodal Biometrics Test;471
41.5;A Case Study;472
41.5.1;The Environment;472
41.5.2;Two Biometrics Recognition Processes;472
41.5.3;Analysis;473
41.6;Conclusions;474
42;Task Migration in a Pervasive Multimodal Multimedia Computing System for Visually-Impaired Users;476
42.1;Introduction;476
42.2;Related Work;477
42.3;Building a ML Knowledge for Configuration Optimization;478
42.3.1;Machine Learning Training to Build User Preferences;478
42.3.2;Alternative Configuration Spaces;479
42.3.3;Optimizing Configuration of User’s Applications;480
42.3.4;Realizing User Task Through Appropriate Modalities and Media;482
42.4;Design Specification and Scenario Simulations;484
42.4.1;Specification for User’s Task;484
42.4.2;Optimizing User’s Task Configuration;485
42.4.3;Specification for Detecting Suitability of Modality;485
42.4.4;Experimental Results;486
42.5;Conclusion;487
42.6;References;488
43;Minimalist Object Oriented Service DiscoveryProtocol for Wireless Sensor Networks;489
43.1;Introduction;489
43.2;Related Work;490
43.3;picoObjects;492
43.4;Abstract Service Discovery Framework;492
43.4.1;Event Channels;492
43.4.2;Place and Play Environment;493
43.4.3;Properties;494
43.4.4;Basic Interface for Actors;494
43.4.5;Interaction Model for Actors;495
43.4.6;Actor Set-Up;495
43.4.7;Multi-requests;496
43.4.8;Service Lookup;497
43.4.9;Legacy SDP Integration;497
43.5;Experimental Results;498
43.6;Conclusions;499
44;A Novel Data Grid Coherence Protocol Using Pipeline-Based Aggressive Copy Method;501
44.1;Introduction;501
44.2;Related Works;502
44.3;The Pipeline-Based Aggressive Copy;504
44.3.1;Network Architecture;504
44.3.2;Integration with Pipeline Transfer Method;504
44.3.3;Algorithm and Example of PAC;507
44.4;Experimental Results and Analysis;508
44.4.1;Suitable Number of Blocks;508
44.4.2;Comparing the Transmission Speed;510
44.5;Conclusions;511
44.6;References;512
45;A Design of Cooperation Management System to Improve Reliability in Resource Sharing Computing Environment;513
45.1;Introduction;513
45.2;Related Works;514
45.2.1;P2P System;514
45.2.2;Reliability Problem;515
45.3;The Design for Cooperation Management System;515
45.3.1;Definition of Cooperation and Cooperation Group;515
45.3.2;Environment and Structure of Cooperation Management System;516
45.3.3;System Design;517
45.3.4;Comparison with Other System;521
45.4;Conclusion;522
45.5;References;523
46;A Peer-to-Peer Indexing Service for Data Grids;524
46.1;Introduction;524
46.2;Related Work;525
46.3;System Overview;526
46.4;Architecture;527
46.4.1;Distributed Hash Tables;527
46.4.2;Mapping Registration;527
46.4.3;Integration;528
46.4.4;Security Model;529
46.5;Implementation;530
46.6;Performance Measurement;530
46.6.1;Single Node Performance;531
46.6.2;System Performance;531
46.7;Future Work;533
46.8;Conclusion;534
47;A Novel Recovery Approach for Cluster Federations;536
47.1;Introduction;536
47.2;Relevant Data Structures;537
47.3;Working Principle;539
47.4;Algorithm Recovery;544
47.5;Comparison;546
47.6;Conclusion;546
48;SONMAS: A Structured Overlay Network for Multidimensional Attribute Space;548
48.1;Introduction;548
48.2;System Design;549
48.2.1;Attribute-ID-Hybrid Space;549
48.2.2;Space Division and Interconnection Rules;550
48.2.3;Basic Operations;551
48.3;Evaluation;553
48.3.1;Time Efficiency and Traffic Overhead;553
48.3.2;Query Performance;555
48.3.3;Summary;557
48.4;Conclusions;557
48.5;References;558
49;Formal Specification and Implementation of an Environment for Automatic Distribution;560
49.1;Introduction;560
49.2;Architecture;561
49.3;Formal Specification;562
49.3.1;Distributed Program Structure;562
49.3.2;Runtime Elements;564
49.4;Implementation Model and Results;566
49.4.1;Parameter Passing;566
49.4.2;Implementation;567
49.5;Evaluations;569
49.5.1;ReadTest Benchmark;569
49.5.2;Warshall Algorithm;570
49.6;Conclusions;570
49.7;References;571
50;Dynamic Distribution for Data Storage in a P2P Network;572
50.1;Introduction;572
50.2;Us System;573
50.2.1;Failure Correlation and Metapeers;573
50.3;Definitions;574
50.3.1;Notations;574
50.3.2;Data Distribution;574
50.3.3;Local Communication Cost of a Peer;575
50.3.4;Global Communication Cost of a Peer;575
50.3.5;Maximal Communication Cost;575
50.3.6;Problem Formulation;575
50.4;Distributions;576
50.4.1;Random Distribution;576
50.4.2;Asymptotically Optimal Data Distribution;576
50.5;Data Distribution in a Dynamic P2P System;577
50.5.1;Metapeer Distribution;577
50.5.2;Over Metapeer Distribution;578
50.5.3;Intrinsic Cost of Metapeer Distribution;579
50.5.4;Analysis of Metapeer Distribution in a Dynamic Way;581
50.6;Conclusion;582
51;GRAVY: Towards Virtual File System for the Grid;584
51.1;Introduction;584
51.2;Data Access Problems in the Grid;585
51.3;Related Work;586
51.4;GRAVY: Solution for Data Access Problems in the Grid;587
51.5;Architectural Issues;588
51.5.1;Protocol Resolution;588
51.5.2;Naming Management;590
51.5.3;File Access and File Transfer;590
51.6;Experimental Results;591
51.6.1;Support for Multiple Protocols;591
51.6.2;Performance;592
51.7;Conclusion;593
52;A Framework for Dynamic Deployment of Scientific Applications Based on WSRF;596
52.1;Introduction;596
52.2;Related Work;597
52.3;Model Design and Implementation;598
52.3.1;The Model Architecture;599
52.3.2;Service Implementation;600
52.3.3;MDS and Scheduling;602
52.3.4;Security;602
52.4;Evaluation;602
52.4.1;Dynamic Deployment Experiments;603
52.4.2;Grid Resource Experiments;603
52.5;Conclusion and Future Work;604
53;Group-Based Self-organization Grid Architecture;607
53.1;Introduction;607
53.2;Previous Works and Motivation;608
53.3;Architecture Outline;609
53.4;Analytical Model and Analysis;611
53.5;Protocol Operation;613
53.5.1;Recovery Algorithms;615
53.5.2;Protocol Messages;616
53.6;Performance Evaluation;616
53.6.1;Testbed;616
53.6.2;Measurement Results;616
53.7;Conclusions;618
53.8;References;618
54;UR-Tree: An Efficient Index for Uncertain Data in Ubiquitous Sensor Networks;620
54.1;Introduction;620
54.2;Related Works;621
54.3;UR-Tree;622
54.3.1;Index Structure;622
54.3.2;Search Algorithm;624
54.3.3;Insert Algorithm;625
54.3.4;Update Algorithm;626
54.4;Performance Evaluation;627
54.5;Conclusions;629
54.6;References;630
55;ZebraX: A Model for Service Composition with Multiple QoS Constraints;631
55.1;Introduction;631
55.2;Related Work;632
55.3;QoS-Based Service Composition Model;633
55.3.1;Definition of the Concepts;633
55.3.2;Definition of the Composition Model;634
55.3.3;Definition of the Utility Function;636
55.3.4;Weight Computation;637
55.4;Implementation;638
55.5;Validation of the Model;640
55.5.1;Validation of the Composition Model;640
55.5.2;Validation of the Decision Model;641
55.6;Conclusion and Future Work;642
55.7;References;642
56;Middleware Support for Java Applications on Globus-Based Grids;644
56.1;Introduction;644
56.2;Execution Model in Globus-Based Grids;646
56.2.1;LCG/gLite Architecture;646
56.2.2;LCG/gLite Job Submission;647
56.3;Execution Model in SUMA/G;648
56.3.1;SUMA/G Components;648
56.3.2;SUMA/G I/O Subsystem;650
56.3.3;SUMA/G Portal;651
56.3.4;SUMA/G Job Submission;652
56.4;SUMA/G Services for Globus Based Grids;653
56.4.1;Security;653
56.4.2;Resource Control and Administration;654
56.4.3;Execution of a Java Application;654
56.5;Experiments;655
56.5.1;Applications;655
56.5.2;Execution;655
56.5.3;Results;656
56.6;Conclusions;657
57;Component Assignment for Large Distributed Embedded Software Development;659
57.1;Introduction;659
57.2;Software Model;660
57.2.1;Design Model;660
57.2.2;Platform Model;661
57.2.3;Deployment Graph;661
57.3;Component Assignment;662
57.3.1;Assignment of Active Periods of Components;662
57.3.2;Priority Assignment of Messages;663
57.3.3;Component Assignment Algorithm;663
57.4;Experiments Evaluation;668
57.5;Conclusions and Future Work;670
57.6;References;671
58;LDFSA: A Learning-Based Dynamic Framed Slotted ALOHA for Collision Arbitration in Active RFID Systems;672
58.1;Introduction;672
58.2;Related Work for Anti-collision;673
58.2.1;Basic Framed Slotted ALOHA Algorithm;674
58.2.2;Dynamic Framed Slotted ALOHA Algorithm;674
58.3;The Proposed Learning-Based Dynamic Framed Slotted ALOHA (LDFSA) Algorithm;674
58.4;Performance Evaluation of LDFSA;678
58.4.1;Simulation;678
58.4.2;Experiment and Verification;680
58.5;Conclusion and Future Works;681
59;Implementation of OSD Security Framework and Credential Cache;683
59.1;Introduction;683
59.2;OSD Security Framework;684
59.3;OASIS Security Framework and Lcache;685
59.3.1;Implementation of OASIS Security Framework;685
59.3.2;LCache;686
59.4;Conclusion;687
60;SEMU: A Framework of Simulation Environment for Wireless Sensor Networks with Co-simulation Model;689
60.1;Introduction;689
60.2;Related Works;690
60.3;The Framework of SEMU;691
60.4;Co-simulation Model;692
60.5;Evaluation Results;693
60.6;Conclusion;693
60.7;References;694
61;Combining Software Agents and Grid Middleware;695
61.1;Introduction;695
61.2;$DG-ADAJ$ Platform;695
61.3;Agent Brokers Augmenting $DG-ADAJ$;697
61.4;Combining Agent-Brokers and DG-ADAJ;700
61.5;Concluding Remarks;702
62;A Web Service-Based Brokering Service for e-Procurement in Supply Chains;703
62.1;Introduction;703
62.2;BPIMS-WS Architecture;704
62.3;Web Services Discovery, Composition, Monitoring and Management in BPIMS-WS;707
62.4;Related Works and Discussion;708
62.5;Conclusions;709
62.6;References;709
63;A Thin Client Approach to Supporting AdaptiveSession Mobility;711
63.1;Introduction;711
63.2;Related Work in the Area;712
63.3;Design Goals and System Architecture;713
63.3.1;Design Goals;713
63.3.2;System Architecture;713
63.4;Implementing an Adaptive Approach to Session Mobility;714
63.5;Evaluating the Performance of Our Approach;715
63.6;Conclusion and Future Work;717
64;Automatic Execution of Tasks in MiPeG;719
64.1;Introduction;719
64.2;Motivations and Contributions;720
64.2.1;Motivations;720
64.2.2;Our Contribution;721
64.3;Utility Framework;722
64.3.1;Service Architecture;722
64.3.2;Scheduling Algorithms;724
64.4;Conclusions and Future Work;725
64.5;References;725
65;Providing Service-Oriented Abstractions for the Wireless Sensor Grid;727
65.1;Introduction;727
65.2;TinySOA: Service-Oriented Architecture for WSN;728
65.3;Implementation and Tests;730
65.4;Conclusions;731
66;Bio-inspired Grid Information System with Epidemic Tuning;733
66.1;Introduction;733
66.2;Ant-Inspired Reorganization of Grid Resources;734
66.2.1;System Entropy and Pheromone Mechanism;736
66.3;Adaptive Tuning and Epidemic Control;737
66.4;Conclusions;740
66.5;References;740
67;Credibility Assignment in Knowledge Grid Environment;741
67.1;Introduction;741
67.2;Credibility Assignment to Decision Makers;742
67.2.1;Formal Definition;743
67.2.2;Solution of the Equation Set;745
67.3;Comparison;745
67.4;Conclusion;746
67.5;References;746
68;Image Streaming and Recognition for Vehicle Location Tracking Using Mobile Devices;747
68.1;Introduction;747
68.2;System Platform and Spatial Relative Distance;748
68.2.1;Spatial Relative Distance;748
68.2.2;Minimization of Spatial for Information on Location;749
68.3;Extraction and Recognition of License Plate Image;750
68.3.1;Image Expression and PDA;750
68.3.2;Image Recognition Stage;751
68.4;Concluding Remarks;753
68.5;References;754
69;Research on Planning and Deployment Platform for Wireless Sensor Networks;755
69.1;Introduction;755
69.2;The Workflow Analysis of the Platform;756
69.3;The Implementation Framework Design;757
69.3.1;Network Deployment;757
69.3.2;Simulation;757
69.4;Conclusions;759
69.5;References;760
70;Server-Side Parallel Data Reduction and Analysis;761
70.1;Introduction;761
70.2;Background;762
70.3;Overview of SW762
70.3.1;Shell-Script Interface;763
70.3.2;Parallel Execution Engine;764
70.4;Results;764
70.4.1;Test Setup;764
70.4.2;Performance Gain;765
70.4.3;I/O Optimization;765
70.4.4;Summary;765
70.5;Conclusion;766
71;Parallel Edge Detection on a Virtual Hexagonal Structure;768
71.1;Introduction;768
71.2;A Virtual Hexagonal Structure;769
71.3;Edge Detection;770
71.4;Experimental Results;772
71.5;Conclusions;773
71.6;References;773
72;Author Index;774
73;More eBooks at www.ciando.com;0

Leseproben

A High-Performance Virtual Storage System for Taiwan UniGrid (p. 43)
1 Introduction

With the rapid growth of computing power and storage capacity of computers, many researchers and scientists have been concentrated on the development of various Grid systems to efficiently utilize distributed computing and storage resources in recent years.

In Taiwan, a community of educational and research organizations interested in Grid computing technologies founded a Grid computing platform, called Taiwan UniGrid [1]. These organizations contribute their resources of computer clusters for sharing and collaboration. The objective of Taiwan UniGrid is to provide educational and research organizations with a powerful computing platform where they can study Grid-related issues, practice parallel programming on Grid environments and execute computing/data-intensive applications.

As similar to other Grid systems, Taiwan UniGrid consists of three primary portions: Computational Grid, Data Grid and Web Portal. Computational Grid is responsible for managing scattered and heterogeneous computing resources and scheduling the jobs submitted by users. Data Grid is a virtual storage infrastructure that integrates distributed, independently managed data resources and allows users to save and retrieve their data without understanding the configuration of underlying storage resources.

Web Portal, developed by National Tsing Hua University, is a uniform user interface by which Grid users can design workflow, submit jobs, manage data, monitor job and resource status, etc. In this paper, we will present the development of the data management system for Taiwan UniGrid. As the distribution of storage resources and the growth of data size, the needs for efficient Grid data management are continuously increasing.

In these years, many research and scientific organizations have engaged in building data management and storage tools for Grids, such as SDSC SRB (Storage Resource Broker) [2], SciDAC Data Grid Middleware [3], GriPhyN Virtual Data System [4], etc. SRB is a general Data Grid middleware that integrates distributed and heterogeneous storage resources and provides virtualized access interface. It has been a production data management tool and adopted by several Grid projects.

Thus, among these tools, we decide to build our virtual storage system for Taiwan UniGrid based on SRB, while developing additional features that are not well supported by SRB. Before implementing the virtual storage system, we elicited requirements from the user and manager needs. Herein, in additional to the basic Data Grid functions provided by SRB, we identify three main requirements of the current development listed as follows.

• High-performance data transfer:
Since the size of data generated by scientific instruments and Grid applications has grown into the range of Terabytes, large data transfer over the Internet usually leads to a long latency and becomes a bottleneck for job executions. Thus, the need for high-performance data transfer is an important issue in Taiwan UniGrid.

• Data sharing:

Two important concepts of Grids are sharing and collaboration. Grid users, such as scientists and researchers, are accustomed to retrieve data collected by remote scientific instruments, analyze these retrieved data via various analysis tools, and share the analyzed results for further processing. Therefore, how to facilitate Grid users to contribute or get shared data with ease is a crucial requirement in the development of a data management system.