Robust approximate lattice alignment design for K-pairs quasi-static MIMO interference channels with imperfect CSI

Abstract

In this paper, we consider a robust approximate lattice alignment design for K-pairs quasi-static MIMO interference channels. The traditional interference alignment on the signal space is infeasible for K > 3 and perfect channel state information (CSI) is required for all the alignment schemes in the literature, which is impractical in practice. Furthermore, the structured lattice, widely used in the practical communications, can create a structured interference space compared with random Gaussian symbols. However, the existing alignment schemes on the lattice codes either require infinite SNR or require symmetric interference channel coefficients. Furthermore, they all assume perfect CSI knowledge. In this paper, we propose an approximate lattice alignment scheme for general complex interference channels in the general SNR regimes as well as a two-stage decoding algorithm to decode the desired signal from the approximated structured interference space. We derive the achievable data rate based on the approximate lattice alignment scheme and formulated the precoder, equalizer as well as the interference quantization design as a mixed integer and continuous optimization problem. By exploiting specific problem structures and incorporating CSI errors in the design, we derive low complexity robust solutions. We show that the proposed scheme has significant gain compared with the existing baselines.