Observation Space Representation (OSR) is a method of encoding and transmitting GNSS corrections where all error sources affecting satellite observations are combined into single composite corrections per satellite and signal. In OSR, the correction provider aggregates effects from satellite orbits, clocks, ionospheric delays, tropospheric delays, and other sources into ready-to-use correction values that rovers apply directly to their pseudorange and carrier phase measurements.
The OSR approach characterizes traditional RTK correction methods, including both single-base RTK and most Network RTK implementations. In these systems, reference stations observe satellites and compute corrections representing the combined errors at their locations. These corrections are transmitted to rovers (via radio, NTRIP, or other means), which apply them to their own observations with appropriate interpolation or extrapolation for location differences. The rover receives and applies corrections without needing to decompose or understand the individual error sources involved.
OSR corrections are typically formatted using RTCM protocols and delivered via NTRIP for real-time applications. Standard RTCM message types convey pseudorange corrections, carrier phase corrections, and supporting information like station coordinates and antenna parameters. The rover’s positioning engine applies these corrections through observation-level adjustments before computing its position solution. This straightforward application makes OSR compatible with diverse receiver hardware and positions computational burden primarily on the correction provider rather than the rover.
The alternative approach, State Space Representation (SSR), models each error source separately and transmits individual corrections for orbits, clocks, ionosphere, and troposphere. SSR corrections require rovers to reconstruct and apply appropriate corrections for their specific location, demanding more sophisticated receiver processing. However, SSR enables more flexible correction application and potentially better performance over wide areas. Modern precision positioning services may use OSR, SSR, or hybrid approaches depending on coverage requirements, bandwidth constraints, and target receiver capabilities.
