版权所有:内蒙古大学图书馆 技术提供:维普资讯• 智图
内蒙古自治区呼和浩特市赛罕区大学西街235号 邮编: 010021
作者机构:School of Electrical Engineering Beijing Jiaotong University Beijing China Automation School Beijing University of Posts and Telecommunications Beijing China
出 版 物:《Chinese Journal of Electronics》
年 卷 期:2025年第23卷第2期
页 面:311-314页
主 题:Ant colony optimization Scheduling algorithms Heuristic algorithms Distributed databases Load management Stability analysis Scheduling Real-time systems NP-complete problem Research and development
摘 要:Distributed data stream processing system is NP-complete problem to assign tasks to any number of nodes handling the task scheduling. Even for substantially reducing scheduling scale, the problem still cannot be avoided. This paper takes advantage of the classical algorithm (ant colony optimization) of heuristic methods to simulate the global task scheduling problem of distributed system. Rational improvement on ant colony optimization path-finding for the memory and CPU usage of each node achieves load balancing in a short time. It gives the sub-optimal solution of the global task scheduling. The experiments show that the data stream processing system we proposed has good real-time characteristics and stability.