A Low-Complexity Joint AOA and TOA Estimation Method for Positioning with 5G Signals

This paper proposes a novel joint angle-of-arrival (AOA) and path delay (a.k.a. time-of-arrival, TOA) estimation (JADE) method for the fifth generation (5G) signal based positioning system. Conventional methods combine angle and delay into a parameter pair and apply super-resolution methods to the large-dimensional space-time/frequency data matrix. The proposed method reduces the computational overhead on the basis of two observations. Firstly, 5G base stations have unbalanced resolutions in temporal and spatial domains, with temporal resolution much better than spatial resolution, especially for those established for indoor coverage. Therefore, the proposed method employs the iterative adaptive approach (IAA) to resolve multipaths in the TOA domain, then applies the spatial digital beamformer subsequently to obtain AOA estimates. Secondly, based on the Vandermonde structure of the delay signature matrix, fast Fourier transform is employed to accelerate the spectrum computation of the IAA method. Comparing the proposed method with the popular multiple signal classification (MUSIC) based JADE algorithm, the estimation error increases by 72% and 26% in numerical simulation and in real-measured channel data, respectively. However, a reduction of nearly three orders for computational cost is achieved.