Performance analysis of DOA estimation using uniform circular antenna arrays in the threshold region

We extend our investigations into the region between "performance breakdown" (threshold) conditions for subspace-based direction-of-arrival (DOA) estimation techniques and maximum likelihood (ML) DOA estimation for uniform circular antenna arrays. Using a new outlier mitigation technique that can be applied to arbitrary geometry arrays, we demonstrate that subspace-based DOA estimation outliers could be reliably rectified up to the point where erroneous DOA estimates start to generate optimally high likelihood ratios (LRs). This is a manifestation of the onset of a discontinuity in the ML estimation, which cannot be rectified within the ML paradigm.