An Efficient Heuristic for Joint User Association and RRH Clustering in Cloud Radio Access Networks

Abstract

Cloud Radio Access Network (C-RAN) is a promising mobile network architecture that breaks down the conventional base station into two main parts: the Base Band Unit (BBU) and the Remote Radio Head (RRH). In this context, deciding to which RRH users connect is known as the user association problem. Moreover, RRHs may be mapped to a single BBU, achieving statistical multiplexing gain. Deciding what RRHs are grouped together is known as the RRH clustering problem. As these two problems are mutually dependent, we formulate in this paper the joint user association and RRH clustering problem. Our objective is to maximize the network throughput, while reducing the network power consumption. Since our joint problem is NP-hard, we propose to decompose it into two sub-problems: the user association (UA) sub-problem and the RRH clustering (RC) sub-problem. First, a low-complexity heuristic, based on the received SINR, is used to solve the UA sub-problem. Second, a low-complexity heuristic, based on the merge-and-split rules, is introduced to solve the RC sub-problem. These two sub-problems are sequentially and iteratively solved until convergence is reached. We further evaluate the performance of our proposed solution. Simulation results show that our proposed heuristic solution for the RC sub-problem strikes a good compromise between computational complexity and performance, in comparison with the optimal exhaustive search method. Furthermore, our proposed solution for the RC sub-problem outperforms the no-clustering method, where one BBU is exclusively dedicated to each RRH, and the grand coalition method, where all RRHs are attached to a single BBU.