A method for precise point positioning with integer ambiguity resolution using triple-frequency GNSS data

Abstract

This paper proposes a method for precise point positioning with integer ambiguity resolution (PPP-AR) using triple-frequency global navigation satellite systems (GNSS) data. Firstly, an enhanced linear combination is developed for rapid fixing of the extra wide-lane (EWL) and wide lane (WL) ambiguities. This combination has improved performance compared to the Melbourne–Wübbena linear combination, and has 6.7% lower measurement error for the GPS L1/L2 signals, 12.7% lower error for L1/L5, and 0.7% lower error for L2/L5. Comparable improvements were also determined for the Beidou and Galileo constellations. After fixing the EWL/WL ambiguities, a full-rank, triple-frequency carrier-phase-only PPP model is proposed with ionosphere constraints. The probability of AR success rate (Ps) is analysed with the LAMBDA method, using a range of carrier phase and regional ionospheric model (RIM) precisions. Results show that a Ps of 99% is achieved within four epochs of data with carrier phase std = 0.002 m and RIM std = 0.1 total electron content unit (TECU); and within six epochs when RIM std = 0.5 TECU. When the carrier phase std was increased to 0.02 m (depicting high multipath conditions), and with use of a low-precision RIM (std = 0.5 TECU), the proposed method gave significantly improved performance over the method proposed by Li et al (2014 GPS Solut. 18 429–42). The direct estimation of the more challenging narrow-lane (NL) integer ambiguity is analysed by multi-epoch averaging of a proposed geometry-free and ionosphere-free triple-frequency linear combination. Tests with GPS data showed that 65.4% of the NL ambiguities were fixed within 10 min, 90.2% within 20 min, and 95.6% within 30 min.