蚁群路由算法中蚂蚁寻径“卡住”问题的研究.pdfVIP

·2106 · 2008 全国博士生学术论坛电气工程论文集 蚁群路由算法中蚂蚁寻径“卡住”问题的研究 赵 晶 佟为明 哈尔滨工业大学电气工程及自动化学院 黑龙江 哈尔滨 150080 【摘 要】 为解决蚁群路由算法中蚂蚁在网络中寻径时必然遇到的蚂蚁“卡住”在网络节点的问 题，分析了蚁群算法应用于网络路由时必须遵守 的、来自网络本身的限制，指出导致蚂 蚁寻径时“卡住”在网络节点的两个根本原因。 在此基础上，采用一种新的蚁群寻径机 制，通过在中间节点和目的节点上实施双重路径检测和在线移除，消除循环路径，使蚂 蚁在“卡住”时能够快速找到最佳后退节点，尝 试新的探寻方向并最终完成寻径。与现 有的三种典型蚁群路由算法进行比较，仿真验证 了新寻径规则更为有效的最优路径寻径 能力。 【关键词】 蚁群算法 蚁群路由算法 寻径机制 网络路由 网络拓扑结构 A new route searching mechanism of ACR algorithm with solution to the problem of ant being-stuck Zhao Jing Tong Weiming Department of Electrical Engineering ，Harbin Institute of Technology ，Harbin 150080 ，Heilongjiang ，China Abstract ：When ant colony optimization （ACO is applied to route packets around network ，the problem of ant being-stuck in network terminal host during the course of path-searching must be solved. This paper analyzes in details constrains ACO for network routing ，points out the three essential reasons by which the problem is caused. Further more this paper presents a new searching mechanism ，through dual path-checked and removing online mechanism by every immediate node ant-visited and destination host respectively ，ants solve the obstacles of ant getting-stuck and path-recurrence at the same time when they build their routes. Results of numerical simulation abstracted from one real network prove new algorithm can not only settle the problem of ant being stuck ，but also achieve a more effective searching for optimization path. Key words ：ant colony optimization （ACO ）