Stochastic Geometry Analysis of Scalable Cell-Free RAN With Dynamic Association and Deployment
Yunxiang Guo, Dongming Wang, Xinjiang Xia, Ziyang Zhang, Jiamin Li, Pengcheng Zhu, Xiaohu You · IEEE Journal of Selected Topics in Signal Processing · 2025
Cell-free radio access network (CF-RAN) breaks away from the traditional cellular network, forming a scalable wireless access network structure. Based on the conventional cell-free massive multiple input multiple output (CF-mMIMO) system, CF-RAN strategically partitions physical layer functionalities into remote radio unit (RRU), edge distributed unit (EDU) and user-centric distributed unit (UCDU), which enable the CF-mMIMO system to achieve a trade-off between complexity and performance in cooperative transmission. We use scalable full-pilot zero-forcing (FZF) combining/precoding in uplink/downlink and consider the impact of channel estimation error and pilot contamination, the closed-form expressions of uplink/downlink achievable signal-to-interference-noise ratio (SINR) of CF-RAN are given. For both uplink and downlink transmissions, we derive the closed-form achievable rate expressions when channel distribution information (CDI) or channel state information (CSI) is known in signal detection, respectively. Addressing the scalability of CF-RAN, the initial access of user equipment (UE) and dynamic RRU association scheme based on the contention mechanism, multiple RRU-EDU deployment schemes, as well as fractional uplink power control and downlink power allocation is considered. The deployment between RRU and EDU determines the performance of CF-RAN, in which we adopt random deployment, clustering deployment based on k-means algorithm, interleaving deployment based on genetic algorithm (GA), interleaving deployment based on graph coloring algorithm (GCA), respectively. Considering the spatial location randomness of UE and RRU, we model the locations of UE and RRU as two independent binomial point processes (BPP) within a limited area, and derive the expression of user rate coverage probability. Finally, the accuracy of our theoretical results is verified through Monte Carlo simulation.