Trellis Exploration Algorithm Aided Precoder Design in Maximizing MIMO Capacity with Limited Feedback

In this paper, a novel precoder design method is proposed for maximizing MIMO capacity. Specifically, under equal power allocation, we employ the trellis exploration algorithm to design the precoding matrix such that the condition number of the resulting effective channel matrix is minimized. Since the trellis exploration algorithm operates on a finite set of phases, the polyphase precoder is first custom designed at the receiver. Then, a small number of bits indicating phase indices of the corresponding precoding matrix is fed back to the transmitter. We show that for a 2 × 2 MIMO, only 8 feedback bits are necessary to achieve desired performance gain. Simulation results demonstrate that the proposed approach outperforms the wellknown "water-filling" scheme at high SNR in terms of ergodic capacity.