Clustering Wi-Fi Users Based on Activity Patterns Using K-Means Algorithm

Abstract

The rapid development of Wi-Fi networks has become a key pillar in
supporting the internet needs of modern society. However, the increasing
number of users and their diverse activity patterns pose challenges in network
management, especially with regard to bandwidth allocation and service
quality. Variations in activity patterns, such as social media, streaming, and
gaming, create different bandwidth requirements for each user group. This
imbalance in resource utilization can result in degraded quality of service,
especially during peak hours. This research aims to address these challenges
by clustering Wi-Fi users based on their activity patterns using the K-Means
algorithm. The data used includes access time, usage duration, connection
intensity, and user activity type. After going through the analysis process,
users are grouped into several clusters based on the similarity of activity
patterns. The clustering results show significant differences between light,
medium, and heavy users in bandwidth consumption and duration of use. The
results of this study contribute to more efficient Wi-Fi network management,
especially in optimizing bandwidth allocation and supporting data-based
decision-making. With customized management strategies for each user
group, the quality of service can be significantly improved, providing a better
experience for Wi-Fi users.