GNSS Multipath Detection using Normalized Early Prompt Late Area Method in Correlation Domain

Multipath is the dominant error source in global navigation satellite system (GNSS) applications. It is highly associated with geometrical surroundings and cannot be fully eliminated by multi-frequency or multi-satellite measurement combinations. To exclude the range measurement corrupted by multipath, multipath detection plays a critical role in GNSS receivers. A concept of using the correlation values to build a multipath detection metric: normalized early prompt late correlation area (NEPLA) index, is proposed. In our preliminary study, NEPLA was initially accumulated non-coherently to address periodic miss detection issues caused by relative carrier phase rotations. To enhance detection sensitivity, the non-coherent NEPLA is redesigned by incorporating the sign of the late (L) channel, enabling coherent NEPLA accumulation. The analytical detection thresholds are derived theoretically through Neyman-Pearson (NP) lemma to restrict the false alarm rate while maximizing the detection rate. The simulation and experiment are conducted for multipath detection performance validation, showing that multipath detection rates are consistent with theoretical expectations. The comparison between the original non-coherent and modified coherent NEPLA methods are given. The results confirm that the redesigned coherent NEPLA method resolved the periodic miss detection problem and widened the relative delay and C/N0 range that the multipath can be effectively detected, where the detection rate is close to unity for multipath with relative delay 0.2 ∼ 1 chip, relative Doppler over 0.5 Hz when C/N0 decreases to 35 dB-Hz.