Calibrating the error from sensor position uncertainty in TDOA-AOA localization

Abstract

The performance of unknown source localization degrades significantly when random sensor position errors exist. In this paper, a calibration source, whose position is known precisely or imprecisely, is introduced into the hybrid TDOA-AOA localization to alleviate the accuracy loss caused by erroneous sensor positions. The performance improvement, attributed to the use of a calibration source, is derived in terms of Cramér–Rao lower bound (CRLB). To get insight into the use of a calibration source, the optimum placement of a calibration source for TDOA-AOA localization in the presence of sensor position errors is investigated. Two closed-form localization algorithms, both composed of two stages, are proposed for the two different cases. In both algorithms, the sensor position errors are corrected through the calibration source in the first stage, and the location of the unknown source is then estimated from the TDOA-AOA measurements in the second stage. The theoretical analysis shows the proposed algorithms attain the CRLB accuracy when sensor position errors and measurement noises are relatively small. Simulations validate the analytical results and the performance of the proposed algorithms.