Optimal direction finding with partially calibrated arrays

This paper is concerned with the problem of optimal (maximum likelihood) direction of arrival (DOA) estimation in situations where the sensor array is calibrated over only a portion of the DOA space. Situations such as this often arise in airborne direction finding when skywave multipath is present. A parameterization is proposed for partially calibrated arrays (PCAs), and the identifiability of the model is discussed for both uncorrelated and correlated signals. It is shown how the signal and noise subspace fitting algorithms are generalized to handle PCAs, and a detection scheme is proposed for individually determining the number of signals arriving from calibrated and uncalibrated directions. The results of several simulation examples are included to validate the analysis.