Abstract
In the control- and user-plane separation (CUPS) based radio access networks (RANs), control-signaling and data are transmitted by the control base stations (CBSs) and data base stations (DBSs), respectively. However, existing studies usually model the C/U-planes as two separate homogeneous networks, neglecting the dependence among the two planes and users. To address this problem, we analyze the coverage probability, spectrum efficiency (SE) and delay considering the dependent features among CBSs, DBSs, and non-uniformly distributed users based on stochastic geometry. Firstly, we present an analytical model for CUPS, where the DBSs are deployed at user hotspots based on Poisson point processes (PPPs), users are clustered around DBSs based on Poisson cluster processes (PCPs), and CBSs are deployed according to a dependent thinning of locations of DBSs based on Matérn hard-core processes (MHCPs). Secondly, we design novel distance-based fractional frequency reuse (FFR) schemes by exploiting the properties of PCP and MHCP to improve the coverage of cell edge users. Thirdly, we derive the distributions of user downlink rates, which are used to analyze the average queueing delay under M/M/C queueing model. Numerical results are presented to verify the efficiency of the proposed model compared to independently distributed BSs and users, and show the dependent BS deployment could significantly improve the coverage of the network.
Funding Information
  • National Key R&D Program of China (2020YFB1807700)
  • Key-Area Research and Development Program of Guangdong Province (2020B0101120003)
  • Key Research and Development Projects of Shaanxi Province (2022KWZ-09)
  • Postdoctoral Research Program of Shaanxi Province (2023BSHYDZZ100)
  • Postdoctoral Fellowship Program of CPSF (GZC20232058)
  • Higher Education Discipline Innovation Project (B08038)
  • National Natural Science Foundation of China (62071352)
  • Royal Society International Exchanges (IEC\NSFC\181395)

This publication has 56 references indexed in Scilit: