Narrow Results

This chapter investigates service-oriented computing in the context of object-oriented distributed platforms. A platform consists of virtual compute resources, a programming environment allowing for the development of distributed applications, and an operating system to run user programs and to make solving complex user problems easier. Service protocol-oriented architectures are contrasted with service object-oriented architectures (SOOA), and then a metacompute grid based on an SOOA is described and analyzed. A new object-oriented network programming methodology is presented in this chapter. It uses the intuitive metacomputing semantics and the new Triple Command design pattern. The pattern defines how service objects communicate by sending one another a form of service messages called exertions that encapsulate the triplet: data, operations, and control strategy.

This paper deals with the resolution of the Quadratic 3-dimensional Assignment Problem hereafter referred to as Q3AP. Q3AP is an extension of the well-known Quadratic Assignment Problem (QAP) and of the Axial 3-Assignment Problem (A3AP). It finds its application amongst others in Hybrid Automatic Repeat reQuest (HARQ) error-control mechanism used in wireless communication systems. This problem is computationally NP-hard. As far as we know, the largest Q3AP instance size solved to optimality is 13 whereas practical Q3AP instance size can be of 8, 16, 32 or 64. Sequential exact methods such branch-and-bound or sequential metaheuristics are therefore not suited to solve large size instances for the excessive needed computation time. In this paper, we propose parallel hybrid genetic-based metaheuristics for solving the Q3AP. The parallelism in our methods is of two hierarchical levels. The first level is an insular model where a fixed number of genetic algorithms (GA) evolve independently on separate islands and periodically exchange genetic material. The second level is a parallel transformation of individuals in each GA. Implementation has been done using ParadisEO framework, and the experiments have been performed on GRID5000, the French nation-wide computational grid. The experimental results produced by our method were confronted with those reported in the literature. The optimum or the best so far known solutions have been reached in a reasonable computation time.

The splitting of Quasi-Monte Carlo (QMC) point sequences into interleaved substreams has been suggested to raise the speed of distributed numerical integration and to lower the traffic on the network. The usefulness of this approach in GRID environments is discussed. After specifying requirements for using QMC techniques in GRID environments in general we review and evaluate the proposals made in literature so far. In numerical integration experiments we investigate the quality of single leaped QMC point sequence substreams, comparing the respective properties of Sobol', Halton, Faure, Niederreiter-Xing, and Zinterhof sequences in detail. Numerical integration results obtained on a distributed system show that leaping sensitivity varies tremendously among the different sequences and we provide examples of deteriorated results caused by leaping effects, especially in heterogeneous settings which would be expected in GRID environments.

One of the fundamental issues in Grid decentralised resource discovery services is high communication overheads that affect the Grid system’s performance significantly. The rationale is that Grid resources are geographically distributed across the world through a wide area network under various virtual organisations. To address the issue, a significant amount of effort has been made by proposing various decentralised overlay algorithms with semantic solutions. Current Grid literature reveals that when semantic features are added into discovery services, the probability of finding resources is enhanced and communication overheads could be better. However, most of the existing decentralised resource discovery models utilise a domain-based semantic ontology with First Come First Serve (FCFS) basis scheduling for allocating Grid resources that can cause job rejection at run time and can pick resources that are far from the user nodes. As a result, communication overheads of the models are affected as the proximity criterion is not being considered in the selection process. To overcome these issues and enhance the application performance, we propose a Unification of Proximity and Semantic similarity for Appropriate Resource Selection (UPSARS) algorithm in a decentralised resource discovery model by using a sub-domain ontology structure for Grid computing environments. The purpose of this unification is to get optimised resources for user jobs (Gridlets) so that Grid brokers could select optimum resources in terms of proximity with high semantic relevancy. The algorithm considers both semantic and proximity criteria and selects the nearby nodes resources and reduces the communication overheads in terms of proximity and latency. We design and implement the model using the GridSim and the FreePastry simulation and modelling toolkits. The experimental results provide promising outcomes to reduce communication overheads and enhance resource allocation performance.

In industrial, scientific and commercial fields, geographically dispersed large data sets are necessary to be analyzed to get a lot of useful information by using distributed data mining (DDM). However, DDM now confronts many new challenges such as seamless-cooperation which can not easily be solved on current distributed computing environment. As the latest development of distributed computing technology, the gird computing may play a key role in solving these challenges. In this paper, the shortcoming of current existing DDM systems is showed, and a service-oriented architecture of DDM on the grid is proposed. The mining algorithm, distributed data sets and computing power are abstracted as web service resource (WS-Resource), which can cooperate to perform DDM as required dynamically. In addition, the paper describes the mining process on the proposed framework in detail.

The necessity of making a robust system to the uncertainties of the different measures taken from the computing infrastructure is one of Grid Computing research challenges. Particularly, the appropriation of relevant information to tasks scheduling from a set of heterogeneous sources and dynamic behavior has been the focus of a significant studies number. Considering this background, this article aims to propose a task scheduling module in Grid Computing with uncertainty regime decision. The proposed module, named fGrid, uses Fuzzy Logic to treat uncertainties. Simulations to evaluate the fGrid were executed with SimGrid simulator use. In addition to promising results, the use of a standard scale with linguistic variables facilitates the manipulation of both the priority list and the scaling, which contributes to ease maintenance.