Multipath is a significant error source in GNSS positioning that occurs when satellite signals reach the receiver antenna via indirect paths after reflecting off nearby surfaces such as buildings, vehicles, terrain, or water. These reflected signals arrive later than the direct line-of-sight signal and can interfere with it, causing the receiver to miscalculate the apparent distance to the satellite and introducing position errors that are difficult to model or predict.
The mechanism of multipath error depends on the interaction between direct and reflected signals at the receiver antenna. When both signals are present simultaneously, they combine constructively or destructively depending on their relative phases, distorting the correlation peak used to measure signal timing. This distortion biases the pseudorange measurement, with potential errors ranging from centimeters to several meters depending on the environment, antenna design, and receiver signal processing. Carrier phase measurements are also affected, though typically by smaller amounts (centimeters) due to the shorter wavelength involved.
Several factors influence multipath severity. Environment is paramount, open fields experience minimal multipath while urban canyons with reflective glass and metal buildings create severe multipath conditions. Signal bandwidth affects susceptibility, wider bandwidth signals (like L5 with 10.23 MHz chipping rate) can better distinguish direct from reflected signals compared to narrow bandwidth signals (like L1 C/A with 1.023 MHz). Satellite elevation matters because low-elevation signals are more likely to reflect off ground and horizontal surfaces. Antenna design, including the use of choke rings and ground planes, can suppress multipath from below-horizon reflections.
Receivers employ various multipath mitigation techniques with varying effectiveness. Narrow correlator spacing and multipath estimation algorithms can reduce code-phase multipath by 50% or more in many environments. Carrier smoothing of code observations averages out short-term multipath variations. High-rate measurements allow detection of multipath through its characteristic oscillation patterns. For surveying applications, extended observation periods enable multipath averaging since reflection geometry changes as satellites move. Despite these techniques, multipath remains one of the most challenging error sources in GNSS, particularly for kinematic applications in reflective environments.
