The network traffic across multiple Wi-Fi access points (AP’s) in high-density environments is unevenly distributed leading to an imbalance of load on these APs. Such load imbalance usually results in less responsive use of internet applications, low bandwidth distributed across the clients, along other network issues that come with increased load on the APs. Hence, there comes a need to optimally balance the load across the APs to reduce any network congestion over an area and to increase the aggregate bandwidth provided to the clients.One solution of the many to solve this problem would be to develop an algorithm that would allow the user equipment’s (UE’s) to decide to associate to AP with the strongest signal by analysing the load on all the APs that are distributed across an area. Reinforcement learning (RL), which is a training method based on reward and action is a technique that has recently been in use in the networking domain as it allows the networking entities to make decisions locally to enhance the networking performance. This paper introduces an approach by incorporating reinforcement learning which takes into consideration the strength and load of APs distributed over an area to decide on which AP to connect hence increasing the network performance.
Access Point Load Aware User Association Using Reinforcement Learning
2022-08-26
1124148 byte
Conference paper
Electronic Resource
English
User Association for Load Balancing in Vehicular Networks: An Online Reinforcement Learning Approach
Online Contents | 2017
|