Abstract: Millimeter-wave communication systems have a large number of available frequency bands. In order to support multiple streams and reduce power consumption and cost, a hybrid precoder design method based on multi-beam training and iterative updating is proposed for millimeter-wave channels. In order to maximize the spectrum efficiency, the proposed method first utilize multi-beam trainings to estimate the angle of arrivals of the dominant paths of the channels and form the alternative beam collection. Then, the analog precoder is iteratively updated to find the optimal paths of the user channels. Furthermore, based on the determined equivalent channel, the digital precoder is determined by establishing the convex optimal mathematical model and solving the objective function on the condition that the spectrum efficiency should exceed the given threshold. The numerical and experimental results show that the spectrum efficiency of the proposed method is superior to those of existing hybrid precoding algorithms in highly correlated channels.

Key words: wireless millimeter-wave channel, hybrid precoding, iterative beam trainings, equivalent channel, optimization design