基于神经网络和蚁群算法融合的QoS组播路由求解

发布时间：2018-06-13 00:14

本文选题：组播路由 + 神经网络　；参考：《计算机测量与控制》2013年12期

【摘要】：针对现有的求解多约束QoS组播路由中存在的问题,结合BP神经网络局部搜索的优势和蚁群算法全局搜索的优势的特点,进行QoS组播路由算法的设计,提出了一种新型的NNAC算法;该算法通过BP神经网络寻找路径的更优解,改善了QoS组播路由路径寻找的方法;通过实验仿真表明,NNAC算法得到最优组播树的总延时为35,总代价费用为21,在完成150个度约束组播路由路径时,NNAC算法在进行最优组播树的寻找成功率上高于AC算法,同时该算法还克服了AC算法易陷入局部最小点的不足。
[Abstract]:Aiming at the existing problems in solving multi-constrained QoS multicast routing, combining the advantages of local search of BP neural network and the advantages of ant colony algorithm in global search, the QoS multicast routing algorithm is designed, and a new NNAC algorithm is proposed. The algorithm improves the QoS multicast routing path finding method by searching for a better solution of the path through BP neural network. The experimental results show that the total delay of the optimal multicast tree is 35, and the total cost is 21. The NNAC algorithm is more successful than AC algorithm in finding the optimal multicast tree when it completes 150 constrained multicast routing paths. At the same time, the algorithm overcomes the shortcoming that AC algorithm is easy to fall into local minimum point.
【作者单位】：山西大学商务学院信息学院;新乡学院计算机与信息工程学院;
【基金】：河南省科技厅基础与前沿项目(132300410204)
【分类号】：TP393.02

【参考文献】