Benefiting from the scientific achievements in recent years, autonomous unmanned aircraft systems (UAS) are rapidly maturing for use in various missions. The attempt to deploy UAS as an extension of the transportation system draws growing attention. One of the key challenges of deploying UAS traffic is to guarantee the safety of UAS flights, which becomes a focus of related research. The tracking system provides the location of airborne flights to support various UAS traffic management (UTM) technologies. Therefore the standardization of tracking system performance is crucial. In this study, a simulation-based study is carried out to analyze the impact of tracking performance on the safety of UTM flights. The indicators of the tracking system are reviewed and summarized. Flight safety is formulated as the probability of an accident, which leads to a design of a pair-wise aircraft encounter scenario. A UTM simulator is established to enable aircraft trajectories generation. Fast-time Monte Carlo simulations are performed with operational uncertainties implemented, which estimates the probability of safety violations. The results of sensitivity analysis on tracking performance indicators show that the sizing of the safety protection boundary affects the performance of the tracking system. In addition, a larger latency of tracking leads to an increase in the lag-time between violation and detection, which potentially affects successful deconfliction.