Routing decision method based on deep reinforcement learning under SDN framework and device thereof

Abstract

The embodiment of the invention provides a routing decision method based on deep reinforcement learning under an SDN framework and a device thereof. The method is applied to SDN controllers and comprises the following steps: obtaining real-time traffic information in a network; determining priority of each traffic; and inputting the real-time traffic information into a pre-trained deep Q network DQN and successively determining routing of each traffic according to priority level sequence of each traffic. The routing decision method based on deep reinforcement learning under an SDN framework and the device thereof are capable of achieving load balance of the networks in the networks of various topological structures, reducing the occurrence of network congestion and achieving optimization of routing policies in the network environment with highly dynamic change of the network traffic.

Description

technical field [0001] The invention relates to the field of communication technology, in particular to a routing decision method and device based on deep reinforcement learning under the SDN architecture. Background technique [0002] For a long time, congestion avoidance and route optimization have been important research topics in traffic engineering in modern communication networks. With the rapid growth of the number of users and network scale, the network structure is becoming more and more complex, and network congestion and route optimization are facing more and more challenges. [0003] The highly dynamically changing traffic services and unevenly distributed traffic density in the network are the main causes of network congestion. In order to solve the network congestion, the common solutions mainly include: performing multi-path distribution on the traffic that may cause the network congestion to prevent the excessive concentration of the load caused by the traff...

AI Technical Summary

Problems solved by technology

And because it is difficult to reduce or avoid network congestion, in a network environment where network traffic is highly dynamic, the routing strategy based on ECMP technology cannot achieve optimal

Images