Skip to Content

Distributed Resource Allocation for Contributory Systems

This chapter introduces the context of the research themes explored in this thesis. It starts with the fundamental motivations behind decentralised and coordinated organisation of distributed systems; including resource allocation and resource management systems.

The chapter thereafter provides discussion on the thesis outline and contributions. It ends with a summary of the published materials that were partially or fully utilised for compiling the thesis.

Contents

Acknowledgments
1 Introduction

1.1 Overview
1.2 Resource Allocation
1.3 Contributions
1.4 Outline
1.5 Publication Record
2 Resource Management and Allocation
2.1 Sate of the art in Contributory Systems

    2.1.1 Collaborative Grids
    2.1.2 Internet Volunteer Grids

2.2 Related work on Resource allocation frameworks

    2.2.1 Auction markets for single type of resource
    2.2.2 Auction markets for multiple types of resources

2.3 Related work on architecture of computational resource markets
2.4 Lessons Learnt
3 Computational Resources
3.1 Introduction
3.2 Resource Properties
3.3 Matchmaking and description of resources
3.4 Ontologies for resource description

    3.4.1 How ontologies meet our requirements?
    3.4.2 Representation of traded resources

3.5 From semantics to a language
3.6 A Formal Bidding Language
3.7 Related Work on Bidding Languages
3.8 A Tree Based Bidding Specification

    3.8.1 Leaf-Node Specification
    3.8.2 Implementation

3.9 Workflow specification
3.10 Support for multiple auction formats

    3.10.1 Bid decomposition
    3.10.2 Integration

3.11 Evaluation
3.12 Conclusions
4 Architectural Approaches for Resource Allocation
4.1 Introduction
4.2 Requirements
4.3 Dynamic Market Deployment for Decentralized Resource Allocation

    4.3.1 Scenario
    4.3.2 Architecture
    4.3.3 Trading process
    4.3.4 Accessing the resources

4.4 Markets in DyMRA
4.5 Configurable Auction Server

    4.5.1 Auction Server
    4.5.2 Architecture
    4.5.3 Workflow and Control
    4.5.4 K-DA Mechanism
    4.5.5 Deployment, Configuration and Execution

4.6 Implementation
4.7 Conclusions
5 Multi Lane Double Auction
5.1 Introduction
5.2 Requirements
5.3 Objective
5.4 The Multi-Lane Double Auction

    5.4.1 Preciseness vs Impreciseness
    5.4.2 MLDA structure and general operations
    5.4.3 How does a double auction work?
    5.4.4 MLDA operations

5.5 MLDA algorithm

    5.5.1 Bid Insertion
    5.5.2 Ask insertion
    5.5.3 Clearing

5.6 Implementation and Experiments

    5.6.1 Experiment A: Economical Efficiency
    5.6.2 Results Analysis
    5.6.3 Experiment A.2 Results
    5.6.4 Experiment A.3 Results
    5.6.5 Experiment B: Computational Efficiency
    5.6.6 Experiment C: Scale Sensibility
    5.6.7 Experiment D: Price per time slot
    5.6.8 Experiment D.1 Results
    5.6.9 Experiment D.2 Results

5.7 Experiment E: Memory Usage

    5.7.1 Experiment E Results

5.8 Conclusions
6 Conclusions and Future Work
6.1 Conclusions
6.2 Future Work

Download
Distributed Resource Allocation for Contributory Systems