Dynamic 5g Throughput Prediction Framework Using Hybrid Machine Learning Approaches

Accurate and dynamic throughput prediction is critical for optimizing network performance and resource allo- cation in 5G networks. This research presents a hybrid machine learning (ML) framework for real-time 5G throughput prediction, integrating multiple ML models to enhance accuracy and adaptability. The proposed architecture employs Random Forest (RF) for feature selection, Boost for boosting predictive performance, and Long Short-Term Memory (LSTM) to capture temporal dependencies in network traffic. By leveraging a diverse set of features, including network traffic patterns, environmental conditions, and user mobility, the model continuously adapts to evolving network scenari- os.Experimental evaluations demonstrate that the hybrid model significantly outperforms standalone ML models in terms of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R2 score, ensuring more reliable throughput es- timation. The framework's adaptability allows network operators to optimize resource utilization efficiently, leading to improved quality of service and user experience. This study highlights the potential of hybrid ML models in tackling real- time challenges in 5G network performance prediction, offering a scalable and robust solution. By combining multiple ML techniques, this approach provides enhanced predictive accuracy, making it a valuable tool for next-generation wireless communication systems.