A Transceiver Design for Limited feedback AF MIMO Relay System with MMSE-SIC Receiver

Precoded multiple-input multiple-output relay (MIMO relay) systems are able to not only extend the linking coverage, but improve the link quality in a cellular system. Generally, the precoding is conducted by either feeding back the channel state information (CSI) or the precoder to the transmitter. However, the perfect feedback is almost impossible due to the feedback overhead. Practically, the limited feedback mechanisms are considered, where only a quantized precoder is fed back to the transmitter. Therefore, how to select a proper quantized precoder (codeword) is a crucial issue. Comparing with conventional MIMO systems, the codeword selection is more complicated in MIMO relay systems due to the additional relay link and the far-end CSI problem. In this paper, we will study three limited feedback mechanisms, categorized by centralized and decentralized schemes, for the amplify-and-forward (AF) MIMO relay system with minimum mean-squared-error successive interference cancellation (MMSE-SIC) receivers. For the centralized codeword selection (CCS) scheme, we propose two selection criteria by minimizing the chordal distance and maximizing the worst-case post signal-to-interference-plus-noise ratio (SINR) criterion. To reduce the overhead burden of exchanging CSI and the computational complexity, we further propose a decentralized codeword selection (DCS) method. Simulation results show that the proposed DCS scheme can approach the proposed CCS schemes with much lower computational complexity and less CSI exchange overhead.