Related Experiment Video
Updated: Jul 13, 2026
08:31
Place and Response Learning in the Open-field Tower Maze
Published on: October 28, 2015
10.4K
A novel and effective method for solving the router nodes placement in wireless mesh networks using reinforcement
Le Huu Binh1, Thuy-Van T Duong2
1Faculty of Information Technology, University of Sciences, Hue University, Hue City, Vietnam.
Plos One
|April 10, 2024
Summary
This study introduces a novel reinforcement learning (RL) approach for optimal router node placement (RNP) in wireless mesh networks (WMNs). The RL method significantly enhances network connectivity, outperforming existing techniques.
Area of Science:
- Computer Science
- Network Engineering
Background:
- Router node placement (RNP) is a critical challenge in wireless mesh networks (WMNs), classified as an NP-hard problem.
- Conventional algorithms struggle with the complexity of RNP, especially in high-density, wide-area WMNs, necessitating advanced solutions.
Purpose of the Study:
- To develop a more effective method for solving the RNP problem in WMNs.
- To apply reinforcement learning (RL) as a novel strategy for optimizing router node placement.
Main Methods:
- The RNP problem is modeled as a reinforcement learning (RL) problem.
- Key RL components are defined: environment (network system), agent (routers), action (coordinate adjustment), and reward (connectivity).
Main Results:
- The proposed RL method successfully addresses the RNP problem.
- Experimental results demonstrate a significant increase in network connectivity, up to 22.73%, compared to state-of-the-art methods.
Conclusions:
- Reinforcement learning offers a promising and effective solution for the complex router node placement problem in WMNs.
- This work represents the first application of RL to RNP, achieving superior network connectivity improvements.
