Joint Pre/Post-processing Design for Large Millimeter Wave Hybrid Spatial Processing Systems

In order to support high-data-rate wireless communication links, millimeter wave (mmWave) systems need to overcome considerable propagation attenuation. In mmWave systems, the small wavelength enables pre-processing exploiting large antenna arrays to provide the required gain. Generally, in traditional microwave systems, pre-processing is done at the digital baseband. However, the cost and power consumption of a radio frequency (RF) chain, which carries out translation between RF and digital baseband, is too high. It is impossible to afford one for each antenna element. This hardware limitation places additional constraints on pre-processing design. In this paper, we use hybrid spatial processing architecture with a lower number of RF chains than the antenna elements. We propose a joint design of digital baseband pre-processing and post-processing based on a weighted minimum mean-squared error (MMSE) criterion subject to the transmit power constraint. Then, the optimization in the RF domain is specified into three criteria according to various error weights. For ease of hardware implementation, we develop a lower complexity transceiver in which pre-processing in RF domain after upconversion is implemented merely using analog phase shifters. Finally, we evaluate our proposed scheme by means of simulation.