UE-centric clustering and resource allocation for practical two-tier heterogeneous cellular networks

The heterogeneous cellular network (HetNet) has emerged as a promising technology for the fifth generation mobile networks that can be used to meet high demand of data rate and better quality of service (QoS) performance. However, the performance of HetNet will depend on how scarce resources such as frequency, time, power and spatial resource are shared among user equipments (UEs) in the system and also how interference is controlled. In this work, we utilise UE-Centric clustering as a tool to effectively determine the interfering base stations (BSs) that cause significant interference to each UE in the network. These interfering BSs together with the serving BSs of these interfered UEs will coordinate and make resource allocation (RA) decisions together to allocate spatial directions to each UE in the network in order to manage interference in the network. We formulate the RA problem as maximizing the weighted sum-rate of the HetNet while fulfilling power, QoS and interference constraints. This optimization problem is non-convex. We readily split the RA problem into two sub-problems: the spatial direction allocation problem and the power allocation problem, respectively. We are able to solve these problems efficiently using SeDumi. Simulation results of our proposed method, show significant improvement.