Performance analysis of Heterogeneous Cloud-Radio Access Networks: A user-centric approach with network scalability

Abstract

Heterogeneous cloud-radio access network (HC-RAN) is a very promising network architecture for future generation wireless communication systems. The centralized computation of HC-RAN provides flexibility to coordinate multi-point (CoMP) transmission of remote radio heads (RRHs). These RRHs can cooperatively form static or dynamic clusters which are network-centric or user-centric to serve the user equipment (UE). To meet the demands of future generation wireless networks, the user-centric approach is gaining much attention, especially in dynamic clustering. In dynamic user-centric CoMP, a set of overlapped clusters can cooperate dynamically to serve UEs. In a large network, the main challenge is the scalability of the computational complexity and fronthaul load with the increase of the number of UEs. To address this, we developed a scalable user-centric HC-RAN by utilizing dynamic cooperation clustering (DCC) framework. We presented an algorithm for joint user association, resource allocation, and cluster formation. We derived the expressions for different combining and precoding vectors to satisfy the scalability condition. We evaluated the performance of the scalable user-centric HC-RAN in terms of the achievable rate per UE at two different levels of cooperation with imperfect channel state information (CSI). The performance of the derived schemes is nearly optimal compared to the unscalable benchmark schemes and the ideal scalable system.