Virtualized network service function chain deployment method based on deep reinforcement learning

Abstract

The invention relates to a virtualized network service function chain deployment method based on deep reinforcement learning, which is used for solving the problem of virtualized network service function deployment under an edge computing background, and belongs to the technical field of edge computing. According to the method, the two problems of virtual function placement and flow routing are solved respectively, deployment of a service function chain with the minimum cost is achieved, and the advantages of deep reinforcement learning can be utilized to meet the flow control requirement changing along with time. According to the method, a neural network is used as a basis for accumulating the reward Q value. In addition, when a sample is input into the neural network, the concept of an experience pool is introduced into deep reinforcement learning. According to the method, the total cost and the end-to-end delay, especially the intermediate processing delay, are considered, and the method is suitable for being applied to a dynamic and complex scene with high requirements for the communication cost and delay of the server.

Description

Technical field [0001] The present invention relates to a deep reinforcement learning and network function virtualization technology, in particular to a virtual network service function chain deployment method based on a deep reinforcement learning algorithm, which is used to solve the virtual network service function deployment problem under the background of edge computing. It belongs to the field of edge computing technology. Background technique [0002] With the advent of the Internet age, various mobile smart terminals have exploded in popularity. All kinds of objects in life are connected to the Internet, which has caused an explosive growth in the amount of network data. According to the forecast of the Internet Data Center (IDC), the total amount of global data in 2020 will exceed 40ZB. Traditional Internet-based cloud computing provides users with network services through the use of a huge resource system on the Internet and uploads data to the cloud computing center ...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to solve the problem of complex and changeable network environment under edge computing, and to provide a deployment method of virtualized network service function chain. The goal is to improve the deployment efficiency and reduce the deployment cost as much as possible.

Images