Statistical Multiplexing Gain Analysis for C-RAN Based on Processing Resource Utilization

Due to the sharing of processing resource, the utilization rate of processing resource in centralized radio access network (C-RAN) is higher than that in distributed radio access network (D-RAN), achieving the so-called statistical multiplexing gain (SMG). In this paper, we focus on analyzing the SMG and the factors that impact SMG, which could provide guidance for the network deployment. Firstly, the SMG model is set up based on the utilization rate of processing resource in C-RAN and D-RAN. Then, by importing the spatial distribution of data traffic, the formula to calculate SMG is derived. Based on the formula, the impacts of the average traffic load and the spatial traffic fluctuation on the SMG are analyzed. Simulations are carried out to verify the correctness of our analysis. Both analysis and simulation results show that the SMG first increases as the average traffic load gets higher, and then decreases when the average traffic load of all cells is higher than a single cell could process. The highest SMG is achieved when the average traffic load is equal to the processing capacity per cell. Moreover, we find that the spatial traffic fluctuation could bring a higher SMG. Applying our model to the real system, the processing resource utilization of C-RAN and D-RAN in the time domain could be evaluated, as well as the SMG.