现有的网络策略还不够成熟, 无法应对由于巨大的流量增长而引起的不断变化的网络条件.随着人工智能领域的不断发展和研究, 深度学习似乎是网络运营商以更加智能和自主的方式控制和管理其网络的可行途径.文中阐述并指出深度学习应用在网络路由路径优化中的必要性, 即基于深度学习的智能路由.通过Tensorflow这一平台构造深度卷积网络, 仿真后与传统的路由策略相比, 证明了基于深度学习的路由方法的有效性.%The existing network strategy is not mature enough to cope with the ever-changing network conditions caused by huge flow growth. With the latest breakthroughs in machine learning or artifical intelligence, deep learning seems to be a viable way for network operators to control and manage their networks in a more intelligent and autonomous way. This paper expounds and points out the necessity of applying deep learning in network routing path optimization, namely intelligent routing based on deep learning. The deep convolution network constructed by Tensorflow platform is compared with the traditional routing strategy through simulation, what proves the effectiveness of the routing method based on depth learning.
展开▼
