Inter-Tier Interference Mitigation in Multi-Antenna HetNets: A Resource Blanking Approach

We study the optimality of almost blank subframes (ABSs) that are used to limit interference by muting macro base stations (BSs) in multi-antenna heterogeneous networks (HetNets). In particular, we formulate a linear program to find the optimal blanking decision that maximizes the average rate of a typical user. A closed-form condition is derived, which depends on the multi-antenna transmission schemes, i.e., the number of antennas used and the number of users served simultaneously in one resource block, as well as network statistics such as the BSs' density in a tier and the coverage probability. This motivates an online implementation that uses instantaneous measurements of the number of BSs and users in the network. We prove that muting macro BSs is generally optimal in single-antenna HetNets but not in multi-antenna networks, where blanking must consider the transmission scheme. Simulation results validate the analysis and show that the throughput significantly decreases when a massive multiple-input-multiple-output (MIMO) macro BS is blanked. This renders the conventional ABSs inefficient to use for fifth-generation (5G) networks which are expected to utilize massive MIMO technology.