Hybrid beamforming for millimeter wave communications
Date
2022-04-29
Authors
Zhan, Jinlong
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Communications over millimeter wave (mmWave) frequencies is a key component
of the fifth generation (5G) cellular networks due to the large bandwidth available
at mmWave bands. Thanks to the short wavelength of mmWave bands, large antenna
arrays (32 to 256 elements are common) can be mounted at the transceivers.
The array sizes are typical of a massive MIMO communication system, which makes
fully digital beamforming difficult to implement due to high power consumption and
hardware cost. This motivates the development of hybrid beamforming due to its
versatile tradeoff between implementation cost (including hardware cost and power
consumption) and system performance. However, due to the non-convex constraints
on hardware (phase shifters), finding the global optima for hybrid beamforming design
is often intractable. In this thesis, we focus on hybrid beamforming design for
mmWave cellular communications both narrowband and wideband scenarios are considered.
Starting from narrowband SU-MIMO mmWave communications, we propose a
Gram-Schmidt orthogonalization (GSO) aided hybrid precoding algorithm to reduce
computation complexity. GSO is a recursive process that depends on the order in
which the matrix columns are selected. A heuristic solution to the order of column
selection is suggested according to the array response vector along which the full
digital precoder has the maximum projection. The proposed algorithm, not only constrained to uniform linear arrays (ULAs), can avoid the matrix inversion in designing
the digital precoder compared to the orthogonal matching pursuit (OMP) algorithm.
For the narrowband MU-MIMO mmWave communications, we propose an interference
cancellation (IC) framework on hybrid beamforming design for downlink mmWave multi-user massive MIMO system. Based on the proposed framework, three successive interference cancellation (SIC) aided hybrid beamforming algorithms are proposed to deal with inter-user and intra-user interference. Furthermore, the optimal detection order of data streams is derived according to the post-detection signal-to-interference-
plus-noise ratio (SINR).
When considering wideband MU-MIMO mmWave communications, how to design
a common RF beamformer across all subcarriers becomes the main challenge.
Furthermore, the common RF beamformer in wideband channels leads to the need of
more effective baseband schemes. By adopting a relaxation of the original mutual information and spectral efficiency maximization problems at the transceiver, we design
the radio frequency (RF) precoder and combiner by leveraging the average of the covariance matrices of frequency domain channels, then a SIC aided baseband precoder
and combiner are proposed to eliminate inter-user and intra-user interference
Description
Keywords
5G, mmWave communications, massive MIMO, hybrid beamforming, interference cancellation