Adaptive calibration of randomly perturbed sensor arrays

The problem of calibrating dynamic sensor locations in a two-dimensional array with the aid of far-field sources is considered. The authors do not know the exact sensor locations, but postulate a known, nominal array geometry in which the location of one sensor and the direction to a second sensor are known. The time-varying sensor displacements from the nominal locations are assumed to be small relative to the intersensor spacing. The auxiliary sources have unknown, time-invariant bearings. They measure the time delays from the sources to the sensors, corrupted by additive noise, and apply an extended Kalman filter. Results of simulations show that calibration can be achieved provided there is enough a priori knowledge of the source bearings