Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Cooperative multiagent congestion control for high-speed networks.

Kao-Shing Hwang1, Shun-Wen Tan, Ming-Chang Hsiao

  • 1Electrical Engineering Department, National Chung Cheng University, Chia-Yi 621, Taiwan, ROC. hwang@ccu.edu.tw

IEEE Transactions on Systems, Man, and Cybernetics. Part B, Cybernetics : a Publication of the IEEE Systems, Man, and Cybernetics Society
|April 15, 2005
PubMed
Summary

This study introduces a cooperative multiagent congestion controller (CMCC) for dynamic networks. CMCC adaptively manages network traffic, achieving high throughput and low packet loss, outperforming traditional methods.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Eliminating Primacy Bias in Online Reinforcement Learning by Self-Distillation.

IEEE transactions on neural networks and learning systems·2024
Same author

Semi-Supervised Detection Model Based on Adaptive Ensemble Learning for Medical Images.

IEEE transactions on neural networks and learning systems·2023
Same author

Non-destructive acoustic screening of pineapple ripeness by unsupervised machine learning and Wavelet Kernel methods.

Science progress·2022
Same author

Application of artificial intelligence ensemble learning model in early prediction of atrial fibrillation.

BMC bioinformatics·2021
Same author

A deep learning method for counting white blood cells in bone marrow images.

BMC bioinformatics·2021
Same author

Lateral Transfer Learning for Multiagent Reinforcement Learning.

IEEE transactions on cybernetics·2021

Area of Science:

  • Computer Science
  • Network Engineering

Background:

  • Traditional congestion control methods struggle with dynamic networks due to propagation delays and difficulty in setting accurate thresholds.
  • Reactive approaches often lead to inaccurate rate selections and suboptimal performance.

Purpose of the Study:

  • To develop an adaptive multiagent reinforcement learning method for effective congestion control in high-speed, dynamic networks.
  • To address the limitations of traditional methods in determining congestion thresholds and source rates.

Main Methods:

  • Proposed a cooperative multiagent congestion controller (CMCC) with subsystems for long-term policy evaluation and short-term rate selection.
  • Utilized cooperative reinforcement signals from a fuzzy reward evaluator based on game theory.
  • Implemented learning procedures for adaptive action selection in time-varying environments.

Related Experiment Videos

Main Results:

  • The CMCC effectively regulates source flow, achieving high throughput.
  • The proposed approach significantly reduces packet loss rates.
  • Simulation results demonstrate improved system utilization and decreased packet losses simultaneously.

Conclusions:

  • The adaptive multiagent reinforcement learning method offers a robust solution for network congestion control.
  • CMCC demonstrates superior performance in dynamic network environments compared to traditional methods.
  • The approach enables adaptive control for enhanced network performance and stability.