Adaptive Beamforming with a Partially Calibrated Distributed Array
Sensor arrays distributed over a large area can theoretically provide high resolution and coherent gain against diffuse noise, but are difficult to calibrate in practice. In many cases, however, some elements of the array may be well-calibrated while others are not. This paper leverages the size and resolving power of a large, distributed array while only requiring calibration knowledge on a small portion of the array. In simulation, we show that a 32-element uniform line array (ULA) with as few as four calibrated elements is capable of detecting and localizing a weak source in the presence of a strong interferer while using only the calibrated portion of the array is unable to resolve the target. Finally, we demonstrate the detection of a weak source in the presence of nearby strong interference using a distributed array consisting of 12 circular microphone sub-arrays with uncertain inter-array calibration.