具有个体记忆的蚁群算法与网络QoS路由研究  

Ant Colony Algorithm with Individual Memory and Its Application in QoS Routing

作　　者：王文国[1] 刘洋 

机构地区：[1]曲阜师范大学信息科学与工程学院,山东日照276826

出　　处：《通信技术》2016年第5期563-568,共6页Communications Technology

基　　金：国家人事部高层次留学人员回国工作资助项目(No.200461)~~

摘　　要：QoS路由算法的目的是在网络中找到满足一定带宽、延时、延时抖动和丢包率等约束要求的路由,该问题是NPC问题。受自然蚂蚁启发,假设人工蚂蚁具有短期记忆能力,使其选择路径时可把最近迭代搜索到的解与自己过去搜索的最优最差解进行比较,动态调整其路径选择过程。蚁群信息素的变化则采用较优解路径更新策略,以加快算法的收敛。蚂蚁将根据搜索到解的情况,判断是否陷入局部最优,若是则改变路径信息素量上下限的大小,使算法跳出局部最优。仿真实验证明,改进的蚁群算法在解决QoS路由选择问题时,能够获得比基本蚁群算法与最大-最小蚂蚁系统更优搜索性能。QoS routing aims to find a route satisfying restraint requirements of bandwidth, delay, delay jit- ter, and packet-loss rate in modem network. Artificial ant colony algorithm is one of the effective methods to solve QoS routing problems. Similar to real ants in nature, artificial ants with short term memory could compare current search result with its best/worst paths in the past, and then adjust its selection behavior dynamically. Simulation with Matlab indicates that the modified ant colony algorithm could acquire better searching performances as compared to the basic ant colony algorithm and max-min ant system.

关 键 词：蚁群算法 QOS路由选择 蚂蚁个体记忆 动态选择 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]