P-persistent energy-Aware handover decisions employing rf fingerprint for adaptive-sized heterogeneous cellular networks

The diverse architectural evolution and intensive data explosion in the forthcoming 5G will have severe impacts on providing seamless and robust mobility management. In this paper, P-persistent energy-Aware handover (HO) decision strategies with mobility robustness are proposed both for intra-handover cases while a femto user equipment (FUE) roams into another femto access point (FAP) and for cross-Tier handover cases while a macro user equipment (MUE)/FUE roams into/out of the FAP in a dynamic cell sizing involving macro-femto two-Tier networks. To approximate the densely deployed small-cells, a unique RF fingerprint (RFF)-based localization is employed to enable efficient small-cell detection by an RFF matching mechanism. The prediction of the HO trigger is jointly determined by a P-persistent decision mechanism that formulates the specific HO behaviors when an MUE/FUE roams into (HO-in) and out of (HO-out) a femtocell in terms of the correlated coverage variance and UE trajectory features, whereas the target selection follows a utility function in consideration of the UE traveling time and the achievable throughput. The closed-form stationary probabilities of the proposed VHO/HHO decisions are analyzed by a Semi-Markov-based framework. In addition, an adjustable sensing mechanism with dynamic intervals is proposed when UE is located far from the RFF matching region, which can have a positive influence on reducing unnecessary UE energy consumption. Numerical results are presented for the decision accuracy analyses (too early, too late, and ping-pong HOs), energy-efficiency, and resource utilization of the two-Tier system. The comprehensive evaluations indicate that the proposed scheme can enhance the mobility robustness and enable an optimal trade-off between the energy efficiency (EE) and system capacity while eliminating the architectural impacts caused by the dynamic topology and the dense deployment for the next-generation macro-femto two-Tier networks.