Multi-agent systems for advanced semantic grid computing

网格计算是一种分布式系统，它支持分布式资源和服务的共享、选择和聚合，以解决科学、商业等不同领域的大规模问题。网格计算的目标是提供可靠的方法和安全的环境，以计算和解决基于协作的问题，将资源聚集在一起。为了充分发挥这种新兴技术的潜力，网格计算越来越多地被认为配备了关于资源、服务、客户、提供者等的语义信息。在这一领域的现代研究中，重点不仅是网络化和高性能计算基础设施，而且还包括信息处理和知识支持。因此，语义网格计算作为当前网格的扩展而出现。本研究计划提出使用多智能体系统来发展先进的语义网格应用，例如电子商务和电子科学。这就提出了尚未得到严格解决的新挑战。特别具有挑战性的问题是语义网格应用程序的灵活性、可验证性和安全性的工程设计。本研究的主要目的是提出一种新的形式模型和计算框架，用于指定和开发灵活、可验证和安全的网格计算。这对电子申请来说是至关重要的。该方法基于多智能体技术的使用，特别是逻辑和基于论证的智能体，以及适用于网格体系结构的扩展验证和安全技术。在体系结构级别，我们建议研究面向服务的网格视图。这种观点基于各种实体根据各种形式的合同相互提供服务的概念。最终，这个想法是在开发系统的当前实践与对web服务、多代理系统和语义网格计算的现代研究的复杂见解之间架起一座桥梁。这里的挑战直接解决了语义网格计算的科学和工程基础。

Grid computing is a type of distributed system that enables the sharing, selection, and aggregation of distributed resources and services in order to tackle large-scale problems in different fields: science,  business, etc... Grid computing aims at providing reliable approaches and secure environments to computing and problem solving based on collaboration by bringing resources together. To achieve the full potential of this emerging technology, grid computing is increasingly perceived to be equipped with semantic information about resources, services, customers, providers, etc... In the modern research into this domain, the focus is not only on networked and high performance computing infrastructure, but also on information handling and support for knowledge. Thus, semantic grid computing emerges as an extension of the current grid.    This research program proposes to use multi-agent systems to develop advanced semantic grid applications such as e-business and e-science. This presents new challenges that are not yet rigorously addressed. Particularly challenging problems are the engineering of flexibility, verifiability, and security of semantic grid applications. The main objective of this research is to propose a new formal model and a computational framework for specifying and developing flexible, verifiable, and secure grid computing. This is of paramount importance for e-applications. The methodology is based on the use of multi-agent techniques particularly, logic and argumentation-based agents, and extended verification and security techniques adapted to grid architecture. At the architectural level, we propose to investigate a service-oriented view of the grid. This view is based upon the notion of various entities providing services to one another under various forms of contract. The idea, ultimately, is to bridge the chasm between current practice in developing systems and the sophisticated insights of the modern research into web services, multi-agent systems, and semantic grid computing. The challenges motivated here squarely addresses the scientific and engineering foundations of semantic-grid computing.