RVQ-Based Beamforming Design with MMSE Criterion in MISO Interference Channels

This paper proposes robust beamforming designs for a multiple-input single-output (MISO) interference channel (IC). Considering the random vector quantization (RVQ) feedback mechanism, we first derive the closed-form expression for the mean-squared error (MSE) metric by averaging over the noise, the quantization error, and the channel amplitude. With the derived MSE result, we devise the robust beamformer by minimizing total MSEs and by minimizing the maximum per-user MSE. Since the optimizations of both designs are not convex, we then propose iterative methods to find out the solutions. The simulation results verify the robustness of both designs when the quantization error exists.