Abstract

In the rapidly field of mobile networks, efficient spectrum management is critical to meet the growing demand for data services and optimize resource allocation. Traditional spectrum management techniques often face challenges in handling dynamic and complex network environments. This study addresses these challenges by proposing an ensemble machine learning algorithm combining Support Vector Machine (SVM), Random Forest (RF), and Decision Tree (DT) classifiers for effective spectrum management in mobile networks. The proposed ensemble model leverages the strengths of each individual algorithm, combining SVM’s robustness in high-dimensional spaces, RF’s capability in handling large datasets with higher accuracy, and DT’s efficiency in rule-based decision making. To evaluate the performance of the ensemble model, a dataset representing spectrum usage patterns in a dense urban mobile network environment was utilized. The model was trained to predict spectrum occupancy and allocate resources dynamically to minimize interference and maximize network throughput. The experimental results demonstrate a significant improvement in prediction accuracy and resource allocation efficiency. The ensemble model achieved an accuracy of 95.6%, surpassing individual classifiers—SVM at 92.3%, RF at 93.1%, and DT at 89.7%. Additionally, the ensemble approach reduced network interference by 18% and increased overall throughput by 23% compared to traditional spectrum management methods. The findings suggest that the proposed ensemble machine learning model provides a more accurate and efficient solution for spectrum management in mobile networks, potentially leading to enhanced service quality and network performance.