Uplink Time Synchronization Based on Time Drift Measurements in Non-Terrestrial Networks

Non-Terrestrial Networks (NTNs) expand wireless communication services to remote areas with limited or no terrestrial network coverage. The 3rd Generation Partnership Project (3GPP) initiated NTN development as a Release 15 study item in 2017 and has been working on the NTN features in the 3GPP specifications from Release 17. Satellite-based NTNs are characterized by long propagation delays and high relative speeds between satellites and user equipment (UE). In particular, uplink time synchronization is important for the performance of OFDM-based NTNs due to the sensitivity of OFDM to timing errors. To achieve time synchronization, uplink signals from different UEs should arrive time-aligned at a base station (or satellite) using timing advance (TA). Although UEs are assumed to use Global Navigation Satellite Systems (GNSS) for uplink synchronization, integrating GNSS modules into UEs may result in increased power consumption, as well as higher complexity and cost. Additionally, standardization efforts are required to exchange ephemeris data of satellites. This paper proposes two uplink synchronization methods that do not rely on GNSS capabilities. In the proposed methods, each UE measures the time drift of downlink signals caused by satellite movement. The first method directly adjusts the TA value using the measured time drift. The second method employs the extended Kalman filter (EKF) to estimate the distance between the satellite and the UE, which is then used to adjust the TA value. Both methods show significant improvements over the conventional method, which relies only on the TA commands. The first method achieves high performance when measurement errors are small, while the EKF method demonstrates superior synchronization performance even in the presence of large measurement errors. The EKF method rapidly converges to accurate values of state variables and can provide useful system parameters for network management and optimization based on its system model.