摘要
针对计算机负荷并行处理的高效稳定和最优化问题,提出了一种智能化任务分配算法。该算法智能化地选择各节点状态和启动策略,将总负荷分配到合适的节点,以使得通信和处理开销最小、负荷均衡,从而使并行处理系统的整体性能达到最佳,成本最低。仿真结查表明,该算法对那些包含少量或适度数量的子处理器网络来说,更为有效和稳定。
An intelligent task assignment algorithm is developed for the efficiency, stability and optimization of a computer parallel processing load. The approach distributes total loads to appropriate nodes by selecting intelligently different node stations and starting strategy so that communication and processing expense accesses minimum and loads balance. Thus the property of the parallel processing system is optimum and the cost is lowest. The simulation result shows that the proposed algorithm is more effective and stable for networks containing small or moderate numbers of children processors.
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2008年第11期2270-2273,共4页
Systems Engineering and Electronics
关键词
并行处理技术
智能化
星型网络
算法
parallel processing technology
intelligent
star network
algorithm
