Dynamic Resource Provisioning for 5G Networks: Comparative Analysis of Various Fuzzy Logic Based Approaches for Efficient Resource Allocation

The applications such as online gaming, video chats, and emergency services demand the efficient deployment of resources in 5G networks in a dynamic fashion. Conventional approaches frequently fail to handle fluctuating workloads, which leads to reduced service quality and underuse of resources. To solve this, this paper proposes a hybrid resource allocation approach that combines fuzzy logic with machine learning (ML) models. This work uses Type-1 fuzzy systems, and Neuro-fuzzy systems with regression to understand real-time characteristics for optimal provisioning. Prediction accuracy is improved via machine learning models, such as Linear Regression (LIR), Random Forest Regression (RFR), Support Vector Regression (SVR), Logistic Regression (LOR) and Gradient Boosting Regression (GBR); performance is measured using metrics like Mean Squared Error (MSE), Mean Absolute Error (MAE) and R2 Score. Among these, the Random Forest Regression model achieves the best results, with the lowest Mean Squared Error of 11.601, Mean Absolute Error of 1.326, and an R2 value of 0.867. Through the integration of fuzzy systems and machine learning models, the technique provides a scalable and adaptive solution to the intricacies of 5G networks, guaranteeing dependable, efficient, and flexible resource management across a range of application requirements.