Recently, cloud radio access network (C-RAN) has been widely regarded as a promising architecture for the next generation cellular networks. By separating traditional cellular base station to remote radio head (RRH) and baseband unit (BBU), C-RAN exhibits many advantages such as lower operating expenditure, higher throughput and energy efficiency. In order to understand the potential of C-RAN, much effort has been devoted to analyzing the performance of C-RAN. However, we notice that existing work usually assumes some preknown stochastic characteristics on the distribution of network traffic demands. In this paper, we would like to derive the backlog conditions in the BBU pool so as to understand how C-RAN can satisfy the real-time requirement in general cases without relying on any stochastic assumption. Specially, we build an analysis framework based on network calculus to analyze the baseband processing capacity of BBU pool in C-RAN. Then, we derive the effective bandwidth, based on the aggregated flow from RRHs, as the upper bound of the processing capacity of BBU pool. With the derived effective bandwidth, the upper bound of the number of BBUs can be also obtained accordingly. Numerical results validate the analysis result.
展开▼
