Hybrid MVDR Beamformer for Subarray Architecture with Sparse Recovery and Manifold Optimization Methods

Hybrid beamforming of mixed analog and digital weights has emerged as an important technology in radar, communications, and their integrated platforms. Fully connected hybrid beamforming (F-HBF) reduces power consumption and transceiver complexity by employing a smaller number of radio frequency (RF) chains than the number of antennas. The cost and complexity of the F-HBF can be further reduced with subarray-based HBF (S-HBF), which does not require a splitter, and uses a smaller number of phase shifters than demanded by the F-HBF. This paper proposes a joint optimization of analog and digital beamformers for obtaining minimum variance distortionless (MVDR) beamformer solution. We show that this non-convex joint optimization can be efficiently solved by leveraging the concepts on manifold optimization and sparse recovery. Computer simulations are used to compare the proposed S-HBF approach with full digital beamforming (DBF) and partial DBF, in terms of the beamformer signal-to-interference-and-noise ratio (SINR). The performance results show that, in general, S-HBF is inferior to full DBF, and the manifold optimization performs better than the sparse recovery approach.