Key points are not available for this paper at this time.
User mobility brings about the spatial correlation in user association (or handover) over a period of multi slots. The correlation makes the classic user association methodology, i.e., typical user method in stochastic geometry, invalid. In addition, the impact of signal to interference plus noise ratio (SINR) on handover cannot be overlooked due to the strong interference caused by the dense base station deployment. To address the above issues, we aim to explore a methodology to analyze the multi-slot network performance for a mobile user considering the SINR effect at the user’s location before it moves and the user’s location after it moves. The main idea is to find the correlated association (and interference) regions in different time slots. On this basis, the expression of multi-slot coverage probability, SINR-based horizontal, and vertical handover rate are derived under maximum average received power (MARP) strategy. Simulation results prove the positive correlation in the associated distance as well as the multi-slot coverage probability, and the impact of SINR on horizontal and vertical handover rate over a period of multi slots. Moreover, MARP strategy always outperforms the nearest distance strategy on multi-slot coverage probability at the almost same cost of handovers. Finally, we found that increasing the picocell density, transmission powers, and biasing can lead substantial loads offloading to picocell but make no significant contribution to the overall coverage.
Zhang et al. (2018) studied this question.