This paper addresses the problem of precoding matrix selection in large-scale MIMO cellular systems, where traditional codebook search methods may result in high complexity for the receiver due to the increased codebook size. For this reason, we propose to first compute an unconstrained reference codeword, and then search the codebook for the best approximation of such reference codeword, using low-complexity distance metrics. We consider two possible choices for the distance metric: the chordal distance, previously used in the literature for codebook design, and a newly defined distance function based on the sum of phase differences between the elements of two complex matrices. We show that the proposed method can match or outperform state-of-the-art approaches (based on capacity or mutual information) with significantly lower per-candidate complexity.
Low Complexity Precoder Selection for FD-MIMO Systems
2016-09-01
174848 byte
Conference paper
Electronic Resource
English