Blind Co-Channel Source Separation in Sparse Interferometric Arrays

Abstract

Interferometric arrays used for radio astronomy and passive interception of RF signals typically use a calibration table to derive direction-of-arrival (DoA) information. But there are some applications, such as separation and identification of multiple source mixtures overlapping in frequency, which do not require DoA information. In those cases, blind source separation / blind beamforming techniques can be applied. Yet the irregular and sparse nature of typical interferometric arrays introduce issues with standard blind algorithms. In particular, traditional assumptions of signal independence and identifiability are sometimes invalid. In this paper, we examine the performance of the Joint Approximate Diagonalization of Eigen-matrices (JADE) blind source separation algorithm in the context of interferometric array processing. We show that while the array structure does introduce some degradation at certain directions-of-arrival, performance remains quite good even in angles where underlying partial ambiguities exist.