<![CDATA[Network traffic congestion is a major challenge in computer network management, especially in campus environments with a growing number of users and increasing internet access needs. The use of the network for various academic and operational activities often causes congestion, which impacts the quality of service and user performance in accessing information and collaborating online. Issues such as high data volume, network performance instability, inefficient bandwidth allocation, and lack of congestion prediction systems are the main contributing factors. This research aims to predict network congestion using the Random Forest algorithm. The data used was obtained from MikroTik devices and packet analysis using Wireshark, including information about active users, jitter, and bandwidth. The research methods include data collection, preprocessing, model training, and model performance evaluation. The research results show that the Random Forest algorithm is capable of predicting network congestion. This can provide important insights for network managers in taking steps to optimize network performance and reduce congestion. By providing a quantitative data-based approach, this research is expected to be an effective solution for addressing network congestion issues, particularly in complex environments such as campuses. Keywords: Network congestion, Random Forest, MikroTik, Wireshark, prediction.]]>
Copyrights © 2025
