Data packet routing algorithm based on multi-agent deep reinforcement learning

A multi-agent, reinforcement learning technology, used in data exchange networks, digital transmission systems, biological neural network models, etc.

Active Publication Date: 2021-03-12
FUDAN UNIV +1
View PDF9 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, literature [12] considers a centralized data flow routing strategy, and requires global topology information and traffic demand matrix

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
  • Data packet routing algorithm based on multi-agent deep reinforcement learning
  • Data packet routing algorithm based on multi-agent deep reinforcement learning
  • Data packet routing algorithm based on multi-agent deep reinforcement learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0051] Set the parameters of the example

[0052] Simulation environment: Python;

[0053] Network topology: such as figure 2 shown;

[0054] Data packet transmission interval: 0.3 ~ 1.0ms;

[0055] Data packet distribution ratio: 10% to 90%;

[0056] Experience return visit pool size: 100;

[0057] Learning rate: 0.001.

[0058] A packet routing algorithm based on multi-agent deep reinforcement learning, the specific steps are:

[0059] Step 1: Initialize the experience playback pool of each router, and initialize each neural network randomly.

[0060] Step 2: Router n observes local information d p and E n , collect shared information C n . Synthesize current state s t :{d p ,E n ,C n} and the hidden state h t , select action a according to the ∈-greedy strategy t .

[0061] Step 3: Router n transmits data packet p to corresponding adjacent node v t , while receiving the reward r t . The current state and the hidden state are respectively transformed int...

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 belongs to the technical field of distributed routing, and particularly relates to a data packet routing algorithm based on multi-agent deep reinforcement learning. In order to relieve the congestion condition in a computer network, an end-to-end adaptive routing algorithm is designed by using a multi-agent deep reinforcement learning technology, and each router completes data packetscheduling according to local information, so that the transmission delay of data packets is reduced. According to the method, firstly, a mathematical model of distributed routing is constructed, specific meanings of all elements in reinforcement learning are defined, then a neural network is trained, and finally algorithm performance testing is carried out in a simulation environment. Simulationexperiment results show that introduction of the deep neural network can mine feature information in an input network state, balance between an unblocked path and a shortest path is achieved, and compared with other common routing algorithms, shorter data packet transmission time delay is achieved.

Description

technical field [0001] The invention belongs to the technical field of distributed routing, and in particular relates to a data packet routing algorithm based on multi-agent deep reinforcement learning. Background technique [0002] Packet routing is a very challenging problem in distributed computer networks, especially in wireless networks of service providers that lack centralized control. In order to minimize the transmission delay, each router needs to determine the next hop node to transmit its data packet. The first and foremost feature of packet routing is a fine-grained packet forwarding policy. Network traffic information cannot be shared between adjacent nodes. Existing routing protocols use flooding strategies to maintain a globally consistent routing table (such as DSDV algorithm [1]), or build on-demand traffic-level routing tables (such as AODV algorithm [2]). Packet routing needs to meet the dynamically changing traffic in current communication networks. ...

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 Applications(China)
IPC IPC(8): H04L12/721G06N3/04
CPCH04L45/14H04L45/12G06N3/045
Inventor 徐跃东游新宇李宣洁
Owner FUDAN UNIV
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

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap