Adaptive Digital Beamformer Autocalibration Using an Injected Noise Source

Abstract

This paper presents a wideband closed-loop relative adaptive calibration method for digital beamformers in phased array applications. Using an internally distributed and injected additive white Gaussian noise source, we equalized the channels relative to a selected channel before beamforming. This relative injection method has a size, weight, and power (SWaP) advantage compared with an absolute method where a copy of the illumination waveform is used as the reference signal. Measured data from an existing 22-phase-center elemental digital beamformer array were used to evaluate the relative method against existing absolute methods. The channel response was extracted and the bit error rate was evaluated using a simulation test bench. A parametric conceptual design was also performed to evaluate the relative SWaP between the relative and absolute methods. The results show that the relative method is 17.6 times smaller, uses 1.3 times less power, and is 2.4 times lower in cost than the absolute method. The results also show that the relative calibration method can be as effective as an absolute method, given that the injected waveform has a higher signal-to-noise ratio than the ambient scene.