Global Navigation Satellite System (GNSS) localization solutions have been applied in various transportation systems; the GNSS application proves current GNSS performance meets non-safety relevant requirements at meter level accuracy. Railway signalling system is a safety on demand system, requires the transportation system to be safe when failure occurs. Applying GNSS in signalling system to provide accurate and safe location solely from train-borne sensors, is currently the hot issue in the development of next generation of train control system (NGTC). In that, GNSS accuracy/continuity/reliability/availability/integrity/safety integrity performance characteristics are vital to the smooth adoption of GNSS. Once the performance indicates that the characteristics stated above meet the requirement according to railway safety standards, then GNSS can be applied. When the train moves along the railway track, the train movement is constrained by the railway track which provides certainty of train trajectory and the possibility of environmental scenario perception in advance. Since GNSS itself suffers from the propagation errors, especially for the train movement in railway stations, the buildings near the track bring multipath and signal blockage effects to the GNSS RF signals. The GNSS performances are degraded and will lead to train location uncertainty and large measurement deviations. In these GNSS performance degraded scenarios, to satisfy safety requirements, the use of compensation sensors becomes necessary. The multi-sensor integrated train location determination system (LDS) solution using multi-constellation and multifrequency GNSS and digital track map (DTM) are designed to benefit the advantages of each sensor. LDS provides possibilities for continuous and trustable train location along the train journey, and the auxiliary information provided by IMU and DTM delivers track occupancy identification with ease. The pre-surveyed DTM is normally generated by precise localization methods as real time kinematic (RTK) and post-processing techniques. Then DTM provides sub-meter accuracy of the point-of-interests (POIs), and the connections between POIs are the defined sections. This paper investigates the DTM topologies with the consideration of section relation in XML definitions, these “frozen” information is applied in map-matching algorithm for movement direction detection and prediction. This paper will illustrate the benefit of DTM and GNSS fusion location, heading information integration for track occupancy. The basic vertical distance map matching algorithms provides the projection of the fusion location on the DTM, then the movement direction is cross-checked with the projected location and DTM topology. In the cross-check algorithm, the recursive Bayesian estimation will be applied. With this, the fusion locations will be on the railway track continuously, but the along track accuracy is not promised all along, especially before the train enters and after the train leaves the turnouts (switches). Thus, the heading information from GNSS is applied for turnout ambiguity area tests. The heading angular velocity together with the DTM turnout section connections are added with the map-matching algorithm to minimize the ambiguity areas. To test the performance of the proposed track occupancy method, the data collected in the high-speed railway line between Shenyang and Fuxin is used. The sensor fusion of GNSS and DTM with trajectory determination is as the first step, then the performance using auxiliary heading and topology information are tested in turnouts. The results will show the possibility of train-borne solely localization techniques for complicated railway station scenarios.