Handover Performance Analysis in 5G Ultra-Dense Networks Using Self-optimizing Hysteresis and Time-To-Trigger

Mobility Robustness Optimization (MRO) algorithms are currently designed utilizing a self-optimization (auto-tuning) approach with the aim of enriching 5G and beyond wireless communication networks. Auto-tuning alleviates the time-consuming manual tuning of handover parameters by network operators. In this paper, we present a handover performance evaluation of MRO schemes in a 5G Ultra-dense network (UDN) based on self-optimizing Hysteresis and Time-to-trigger. In the UDN scenario, network performance may severely deteriorate due to the severity of interference and frequent handovers for highly mobile users. The evaluation is based on handover event A3 for UE speeds of up to 140 km/h. Extensive simulations are performed for different UE trajectories, and the simulation results of three self-optimization schemes based on Fuzzy Logic, Regression, and Analytical approaches are analyzed in terms of handover success and area spectral efficiency. The Analytical approach achieved higher handover success (99.84%) followed by Fuzzy Logic (67.10%) and Regression (48.58%). Analysis shows that handover performance does not necessarily correlate with area spectral efficiency performance; there exists a performance trade-off. This implies that advanced self-optimization techniques are required to address the handover challenges posed by cell densification.