Limited-Feedback MU-MIMO Systems with MMSE Precoding Design

Limited feedback is a key technique for conveying channel state information (CSI) back to the base station (BS). However, its reliance on quantization to select the optimum code-word from a predefined codebook results in severe degradation in achievable rate due to quantization error. To address this issue, robust techniques should be developed. In this paper, we first examine an approximation for the second-order statistics of quantized CSI. Based on the proposed approximation, we then propose a novel robust precoding design that minimizes the conditional expectation based mean square error (MSE). Numerical results show that the proposed design significantly improves the achievable rate compared to conventional precoding schemes.