TensorFlow平台深度学习的网络路由路径优化选择研究

Research on optimization selection of network routing path based on deep learning of TensorFlow platform

基于深度学习在提高网络控制和管理过程的自主能力和智能水平方面发挥的重要作用,为进一步优化网络路由路径,文中通过结合运用深度学习和TensorFlow平台的方法完成具体的搭建以及训练过程,通过NS3网络仿真的使用完成最优路径的寻找,完成了深度卷积网络及网络智能路由路径优化选择方案的构建,针对网络的平均时延、信令开销、整体吞吐量的性能进行仿真,验证了该路由方法的有效性。

Based on the important role that deep learning plays in improving the autonomy and intelligence of the network control and management process,in order to further optimize the network routing path,the article uses deep learning and TensorFlow platform methods to complete the specific construction and training process.Through the use of NS3 network simulation to find the optimal path,complete the construction of the deep Convolutional network and the network intelligent routing path optimization selection plan,and simulate and verify the performance of the average network delay,signaling overhead,and overall throughput.This improves the effectiveness of the routing method.