Robust Noise Filtering in Wideband Frequency-Invariant Beamforming with Uniform Circular Arrays

Accurate signal measurement in the spatial-temporal domain is critical, amongst other applications, to the commercial success of UWB-MIMO communication systems and UWB standalone location systems. Antenna arrays can measure the spatial dimension of a signal through beamforming; to this end we chose to implement the uniform circular array in our spatial-temporal channel sounder due to its constant beam pattern around the azimuth angle. In wideband systems such as ours, it is important that the beam pattern also be constant across the band of operation to enable linear methods for multipath extraction or sidelobe suppression. Frequency-invariant beamforming can achieve this but at the expense of greater noise. Most literature treating the practical implementation of frequency-invariant beamforming concentrates on mutual coupling and physical-array imperfection. Rather in this paper we propose a novel filtering technique specific to the characteristic noise in frequency-invariant beamforming. It employs the eigendecomposition followed by least-squares minimization and proves robust even at low signal-to-noise ratios as substantiated through simulations as well as measurements using our sounder.