Precise Direct Position Estimation: Validation Experiments.

Direct position estimation is an alternative GNSS positioning method, which parameterizes the received satellite signals as a function of the target dynamics and solves the position, velocity, and time (PVT) quantities directly from the raw signal, as opposed to using observables. It is also recognized as a one-step positioning method, in contrast to the traditional two-step GNSS positioning method which first estimates the observables and then estimate PVT through those intermediate quantities. It has been shown that DPE outperforms the two-step method in low-SNR scenarios, where the early combination of the signals improve the receiver sensitivity and its exploitation of weak signals. This paper improves the existing DPE approach by modeling the carrier phase components in terms of the position of the receiver. In this new framework, termed Precise DPE (PDPE), a new maximum likelihood estimator is derived based on the proposed signal model. This article present the first results of PDPE using a real raw GNSS signal dataset, which is used to compare PDPE performance against DPE and two-step solutions. Results show that PDPE provides improved precision compared to DPE and, additionally, outperforms the two-step method in precision and sensitivity.