Evolutionary Hierarchical Multi-Criteria Metaheuristics for Scheduling in Large-Scale Grid Systems

丛 书 名：Studies in Computational Intelligence

版本说明：2012

作     者：Joanna Kołodziej 

I S B N：(纸本) 9783642289705 

出 版 社：Springer Berlin Heidelberg 

出 版 年：2012年

主 题 词：Computational intelligence Computational grids (Computer systems) Computer scheduling 

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 

摘      要：One of the most challenging issues in modelling today s large-scale computational systems is to effectively manage highly parametrised distributed environments such as computational grids, clouds, ad hoc networks and P2P networks. Next-generation computational grids must provide a wide range of services and high performance computing infrastructures. Various types of information and data processed in the large-scale dynamic grid environment may be incomplete, imprecise, and fragmented, which complicates the specification of proper evaluation criteria and which affects both the availability of resources and the final collective decisions of users. The complexity of grid architectures and grid management may also contribute towards higher energy consumption. All of these issues necessitate the development of intelligent resource management techniques, which are capable of capturing all of this complexity and optimising meaningful metrics for a wide range of grid applications. This book covers hot topics in the design, administration and management of dynamic grid environments with a special emphasis on the preferences and autonomous decisions of system users, secure access to the processed data and services, and application of green technologies. It features advanced research related to scalable genetic-based heuristic approaches to grid scheduling, whereby new scheduling criteria, such as system reliability, security, and energy consumption are incorporated into a general scheduling model. This book may be a valuable reference for students, researchers, and practitioners who work on or who are interested in joining -- interdisciplinary research efforts in the areas of distributed and evolutionary computation.