TOA error bounds for positioning in 5G new radio networks

This paper proposes a framework for the numerical computation of the Cramér-Rao lower bounds (CRLBs) on the achievable time of arrival (TOA) estimation error variances in fifth generation (5G) New Radio (NR) systems. These theoretical bounds are obtained for channel models including line-of-sight (LOS) and multipath propagation in presence of additive white Gaussian noise (AWGN), with parametrization based on the 3GPP standardization. Because 5G NR physical layer (PHY) simulations are complex and time consuming, a simple CRLB model for TOA estimation is instead here proposed, based on pre-computed tables for certain propagation conditions, SNR levels and transmit signal configuration. The proposed model can be understood as a generalization of the well-known closed-form expression of the CRLB for TOA estimation for pure LOS channels (i.e. with no multipath propagation) to the case of LOS channels with multipath propagation. System-level simulations are then performed in order to validate the proposed CRLB model for TOA estimation and to assess the 5G NR downlink time-difference of arrival (TDOA) positioning capabilities.