Performance Assessment of Massive MIMO Systems for Positioning and Tracking of Vehicles in Open Highways

Abstract: The next generation of mobile networks (5G) is currently being standardized, and massive MIMO (Multiple-Input-Multiple-Output) is a strong candidate to be part of this standard. Other than providing higher data rates and lower latency, high accuracy positioning is also required. In this thesis, we evaluate the achievable performance of positioning using massive MIMO systems in open highway scenarios. Relevant theory from sensor array signal processing and Bayesian filtering is presented, and is used in a simulation environment on large antenna arrays representing massive MIMO base stations. Positioning is done by utilizing the uplink pilot reference signals, where the Direction of Arrival (DOA) of the pilot signal is estimated, and then used for position estimation. Estimation of the DOA is done by both a maximum-likelihood method and by using an Extended Kalman Filter (EKF). A positioning error of less than 8 m is achieved with absolute certainty when the vehicle is less than 300 m from the base station. It is also concluded that this result could be improved by using more sophisticated filtering algorithms.