Data-informed CRLB Derivations for Indoor Emitter Localization

In quantifying the performance of indoor localization techniques, Cramér-Rao Lower Bound (CRLB) is often employed. CRLB derivations typically use simplifying assumptions such as analytically tractable distributions which are further assumed to be identical and independently distributed (i.i.d.). In this paper, several data-informed analytic CRLB’s are derived using non-identical distributions to determine the combined site and system-specific bounds in localizing uncooperative wireless emitters. It is shown that this approach leads to a lower bound compared to the standard i.i.d. approach. From a synchronized radio frequency (RF) sensor network, over three thousand measurements consisting of time-difference-of-arrive, received signal strength, and angle-of-arrival measurements taken across a 2100-sq meter area serve as the basis for single measurement and hybrid CRLB’s.