Wideband Adaptive Beamforming for a Partially-Calibrated Distributed Array

This paper introduces two novel array processing techniques to enable wideband beamforming on sparse arrays without calibrating the full array. First, we develop a new wideband focusing method using parameterized unitary transformations to minimize the rank of the wideband data covariance. This enables us to estimate a target-free covariance and ultimately suppress interference without nulling the target as well. Second, we introduce a new partially-calibrated adaptive beamformer (PC-ABF) which leverages the calibration of just a single sub-array to localize the target after interference has been suppressed. Combined, these techniques blindly enables the source separation capabilities of a full distributed aperture.We show experimental results demonstrating interference suppression and weak target resolution on a distributed array aperture. Using in-air experiments on a 64-element array and having calibration on only 16 elements, our PC-ABF demonstrates high-resolution interference suppression and target localization in a complex indoor multipath environment.