基于混合并行遗传算法的网格资源分配策略

A Grid Resource Allocation Strategy Based on Hybrid Parallel Genetic Algorithm

网格是利用互联网或专用网络将地理上广泛分布的、异构的、动态的资源互联起来实现资源高度共享与集成,为用户提供高性能的计算、管理和服务等功能的一项新技术。但如何将网格这个复杂环境中的资源有效进行管理和调度,是一个NP难问题,同时也是网格技术发展的关键。启发式算法被证明是解决这类问题的有效算法,将两个启发式算法结合起来,充分发挥各自的优势,就能有效解决网格资源分配的问题。因此,利用混合并行遗传算法来解决网格资源分配是可行的。

Grid is a new technology that uses Internet or special network to interlink the distributed, heterogeneous and dynamic resources in the world to realize highly share and integration of resources and to provide users with high performance functions such as computing, managements and services. Heuristic algorithm has proved to be a kind of effective algorithm to solve this kind of problems. To joint two heuristic algorithms can fully exert each advantage, and can effectively solve the problem of allocate grid resources. So, it is feasible to use hybrid parallel genetic algorithm to solve the problem of grid resources allocation.