Abstract

Compared with conventional two-step localization algorithms which are implemented by initially estimating localization parameters and then calculating the target position, the performance of direct position determination (DPD) algorithm is superior in terms of high estimation accuracy and strong resolution capability. DPD algorithms can make full use of the characteristics of signals to improve positioning performance. Using this localization mechanism as basis, this study develops a novel DPD algorithm that profits from the characteristics of orthogonal frequency division multiplexing (OFDM) signals based on the time and angle of arrival. First, the DPD optimization model is constructed based on maximum likelihood criterion and the characteristic of multiple carrier modulation technique of OFDM signals. Then, an extended subspace data fusion-based DPD algorithm is developed by constructing and decomposing the extended covariance matrices. The algorithm fuses the extended subspace data to estimate the target position without calculating the intermediate variables, which is less complex and more efficient. These properties can avoid the limitations of two-step algorithms. The last but important result is a derivation of the closed-form expression of Cramer-Rao lower bound (CRLB) on the position estimation variance for OFDM sources. Simulation results show that, the algorithm proposed in this paper outperforms existing DPD algorithms and conventional two-step localization algorithms. Especially under the condition of low signal-to-noise ratio, its localization accuracy is close to the CRLB.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call