Refinement and derivation of statistical resolution limits for circular or rectilinear correlated sources in CES data models

This paper analyzes the deterministic (DCRB) and stochastic (SCRB) Cramér-Rao bound on direction-of-arrival (DOA) estimation for two equi-powered correlated complex circular or rectilinear sources affected by white noise in different elliptically symmetric (CES) data models. Beginning decomposing these CRBs, into factors depending signal parameters, geometric parameters of array, some new interpretable closed-form expressions are derived particular scenarios. These provide useful insight behavior CRB’s dependence correlation factor. Approximate CRBs small DOA separation also derived. results lead to formulas statistical resolution limit (SRL) based Smith criterion at which an unbiased algorithm can resolve closely-spaced sources. SCRB-derived much less optimistic than those have so far been deduced only from DCRB under assumption known