Interference management approaches for MU-MIMO in shadow fading channels

A MIMO interference network with L users is investigated in this paper. We compare the sum-rate performance of interference management approaches applied to this network. The minimum mean square error (MMSE) algorithm optimizes the beamforming vectors by maximizing the signal-to-interference-plus-noise ratio (SINR). Another approach, minimum leakage interference alignment (ML-IA), minimizes the leakage interference only. We propose an approach for using ML-IA when the links have extra degrees of freedom. In particular, we use them to maximize the desired signal power within the desired signal subspace; we call this ML-IA with Max SNR (ML-IA-MS). Simulation results with different numbers of antennas show that ML-IA-MS has comparable performance in most cases. An additional contribution of this work is that it gives quantitative results on the impact of shadow fading and number of beamforming iterations, which are often omitted in similar studies.