An Optimal Stopping Approach to Cell Selection in 5G Networks

Initial cell search and selection is one of the first few essential steps that a mobile device must perform to access a mobile network. The traditional maximum received power based initial cell selection may result in unbalanced loads across the cells in multi-tier heterogeneous 5G networks with millimeter wave. The objective of this work is to improve the initial cell search and selection scheme to achieve a more balanced load distribution in 5G networks. We propose to augment the existing pure maximum received power based scheme with a new load factor broadcast as part of system information. To study the theoretical performance of the proposed initial cell search and selection scheme, we formulate a throughput optimization problem and solve it by using optimal stopping theory. We characterize the throughput optimal stopping strategy and the attained maximum throughput. We prove that the proposed initial cell search and selection scheme is throughput optimal with a carefully optimized connection threshold. The optimal connection threshold and maximum throughput can be found by solving a fixed point equation. We further develop an iterative algorithm to compute the solution to the fixed point equation and prove the convergence of the proposed iterative algorithm. We also provide simulation results to demonstrate the analytical results and obtain insights. Our analysis and results show that the proposed initial cell search and selection scheme can help achieve a more balanced load distribution in 5G networks.