SDN-Assisted Learning Approach for Data Offloading in 5G HetNets

With the fast proliferation of the smartphone usage, the mobile traffic has far more exceeded than the capacity of the LTE-A (Long Term Evolution-Advanced) networks. Hence, the Mobile Network Operators (MNOs) are looking for alternative opportunities to handle the network traffic instead of deploying cost-incurring classical devices. Mobile data offloading, which refers to diverting traffic from cellular networks to other complementary technologies such as WiFi access points offer to be a promising solution. APs provides better data services due to the small coverage area (100 m) and improves battery life. WiFi offloading, when implemented using Software-Defined Networking (SDN) helps in the dynamic management of a complex Heterogeneous Networks (HetNets). In this paper, we propose a novel SDN-Assisted Learning Approach (SALA) to provide better Quality of Experience (QoE) for both the cell edge users and intensive network users at the LTE-A base station using the unlicensed spectrum of the APs. We then verify our novel SALA framework against simulations based on real-world usage, and offer insight to the expected offloading gains in practice.