Intelligent Handover Management in Ultra-Dense 5G Networks: A Deep Q-Learning-Based Prediction Model

Handover optimization is problematic in ultra-dense 5G networks because of the extremely high number of users and base stations. Unlike other methods, classical optimization approaches are not suitable for such technologies due to their complexity. In this context, a Machine Learning (ML)-based architecture using Deep Neural Networks (DNN) and Reinforcement Learning is proposed for handover and network optimization. The DNN model considers the temporal changes in the state of the network, and reinforcement learning is responsible for policy-based decision-making for proactive intervention, adaptive-driven interference suppression, and spectral efficiency improvement. These findings support the feasibility of the proposed resource allocation technique using ML in wireless communication systems.