Position Estimation in Massive MIMO Systems with Low-Resolution ADCs

In this paper, a mobile station (MS) localization scheme is proposed for millimeter wave (mmWave) massive MIMO systems, in which both base station and MS are equipped with multiple antennas. To reduce the hardware and power cost, low-resolution analog-to-digital converters (ADCs) are employed at the base station (BS). The proposed scheme consists of two stages: channel parameter extraction and position estimation. In the first stage, we adopt the generalized turbo (GTurbo) framework to cope with the nonlinear quantization effect caused by low-resolution ADCs, and adopt the sparse Bayesian learning (SBL) framework to exploit the sparsity in mmWave channels. The combination of GTurbo and SBL enables us to accurately extract the localization-related channel parameters. In the second stage, the MS position is precisely estimated by using the Gauss-Newton algorithm. Numerical results verify that the proposed scheme can achieve the performance close to the Cramér-Rao bounds (CRBs).