Cramer-Rao lower bounds for multi-sensor localization

Abstract

The Cramer-Rao lower bound is used to assess the potential localization accuracy of multiple arrays observing a narrowband moving target. The narrowband signal received by the array is assumed to have only partial temporal coherence, which is modelled by taking the signal to be completely coherent over a data block but with an unknown absolute phase from block to block. The signal may also have unknown phase differences between different sets of sensors. Including these unknown phases in the model enables to account for the spatial incoherence of the signal between arrays. Examples of the localization accuracy possible for multi-array or multi-sensor systems as a function of target position are given.<>