A q-learning-based path selection method for software-defined networks

A software-defined network and path selection technology, applied in the field of communication, can solve problems such as unknown devices and paths, dynamic changes, service requests and network nodes do not correspond one-to-one

Active Publication Date: 2019-07-30
JIANGSU ELECTRIC POWER CO
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In the network, how to choose the best path is very important for the entire network service system; on the other hand, it is not easy to find the path in the software-defined network for two main reasons: first, the software-defined network The service requests in the network are not in one-to-one correspondence with the network nodes, so it is necessary to map the service to the network nodes while finding the path; secondly, the devices and paths in the network may be unknown, and may also change dynamically

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A q-learning-based path selection method for software-defined networks
  • A q-learning-based path selection method for software-defined networks

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0026] Embodiment one: see figure 1 , 2As shown, a path selection method for software-defined networks based on Q-learning, the software-defined network infrastructure layer receives service requests, and the software-defined network controller constructs a virtual network according to the required service components and combination methods, and allocates suitable The network path completes the service request and finally reaches the terminal, and the suitable network path is obtained through the Q learning method in reinforcement learning, and the method steps are:

[0027] (1) Set up several service nodes P on the established virtual network, and each service node is assigned a corresponding bandwidth resource B;

[0028] (2) Classify the received service requests into actions a that can be taken, and try to select each path that can reach the terminal according to the ε-greedy strategy, that is, each action a passes through the corresponding service node P to complete the ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a path selection method for software-defined networks based on Q-learning. The software-defined network infrastructure layer receives service requests, constructs a virtual network, and allocates suitable network paths to complete service requests. It is characterized in that: The suitable network path is acquired through Q-learning method: (1) Set several service nodes P on the established virtual network, and each service node is assigned a corresponding bandwidth resource B; (2) Decompose the received service request into available Action a, according to ε‑greedy, try to select every path that can reach the terminal; (3) record data into a Q value table and update it; (4) find a suitable path according to the record data in the Q value table. The present invention uses the Q learning method to find a forwarding path that is short, consumes less time, occupies less bandwidth resources, and is suitable for network paths of dynamic and complex networks, and satisfies as many other service requests as possible without adjusting the virtual network .

Description

technical field [0001] The invention relates to the field of communication technology, in particular to a Q-learning-based path selection method for a software-defined network, which can find the most suitable service path to satisfy a service request on the basis of an existing virtual network. Background technique [0002] In recent years, people have diversified requirements for the types of information obtained in the network, and the requirements for the quality and security of information obtained in the network have also been continuously improved. The amount of information carried by various networks is rapidly expanding, the scale of the network is constantly expanding, and more and more users, applications, and services are connected to the network. Network construction, expansion, optimization, and security work have become important contents of network construction and maintenance. However, in the face of these complex and changing needs, the original Internet a...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Patents(China)
IPC IPC(8): H04L12/751H04L12/707H04L45/02H04L45/24
CPCH04L45/02H04L45/24
Inventor 景栋盛薛劲松王芳朱斐
Owner JIANGSU ELECTRIC POWER CO
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products