on the cramer-rao lower bound based analysis of direct position determination and doa position finding for co-channel emitter localization

Direct position determination (dpd) is known as an optimal, single-step technique for localizing co-channel signal sources since it processes the data gathered from all the array receiver elements together. in contrast, the commonly used radio location techniques include two independent stages. first, they estimate some initial parameters like direction, time, time-difference, frequency of arrival, etc., or their combination, and second, they localize signal sources using the triangulation of loci generated by the first stage. this disjoint structure leads to the sub-optimality of conventional localization algorithms. in this paper, we compare the location root-mean-square-error lower bounds (lelb) for dpd and position finding by doa (pf-doa) to prove the superiority of dpd over pf-doa, which are commonly used for tactical fields or outdoor applications. moreover, we demonstrate the advantages of dpd for indoor localization applications compared to pf-doa techniques in terms of localization accuracy. we also introduce the single-group-array (sga) structure for dpd in indoor applications and reveal that it outperforms both the pf-doa and dpd with a classical multi-group-array (mga) structure.