Adaptive inter-cell interference cancellation in heterogeneous networks: Making smart use of multiple antennas at base stations

Abstract

In downlink heterogeneous networks (HetNets), inter-cell interference management becomes more complicated and challenging than that in traditional homogeneous networks due to the coexistence of different types of base stations (BSs). Most existing interference management approaches, such as coordinated beamforming and joint transmission, require joint signal processing at a number of, if not all, BSs. The burden thus imposed on the backhaul network may be too heavy for practical systems. In this paper, we therefore propose a low-complexity inter-cell interference cancellation (ICIC) scheme, where each BS independently cancels its interference to a set of users in neighboring cells when transmitting to its own user. In particularly, the proposed ICIC scheme is implemented in two time-scales. In the slow time-scale, each BS is assigned to a set of users intended for ICIC, referred to as a nulling set, according to long-term statistics of the wireless channels. Then, in the fast time-scale, beamforming vectors are chosen at each BS to cancel its interference to the users in its nulling set based on the instantaneous channel state information (CSI). In this way, the overall system throughput is improved through slow adaptive assignment of nulling sets, while real-time signal processing is implemented locally at each BS. To assign nulling sets for BSs, we should carefully examine the trade-off between the gain from ICIC and the loss in the transmission diversity, such that the benefits of multiple antennas at BSs are exploited efficiently. To this end, we formulate the problem of finding the optimal assignment of nulling sets as an integer programming (IP), where the data rate is expressed as a function of the nulling set assignment. With the derived expression of the data rate, we develop a greedy algorithm to solve the IP efficiently. Simulation in a two-tier network shows that the proposed ICIC scheme can balance the data rates in pico and macro cells adaptively such that the total throughput is improved.