A GNSS Anti-Interference Integer Float Solution Based on LAMBDA Integer Transform

GNSS is an important means that can provide high-precision navigation and positioning information for intelligent driving. In complex urban environments, after briefly losing the GNSS signal, it takes initialization time for a vehicle to regain high-precision positioning information. Therefore, shortening the initialization time is an important step in providing real-time continuous navigation and positioning services for intelligent driving. The integer float estimator solution has the advantage of free initialization, which can greatly reduce the convergence time of ambiguity fixing. However, its positioning error may show a sudden increase under poor observation conditions. Aiming at the problem that the integer float estimator may be interfered with, this paper proposes an anti-interference integer float estimator method for GNSS based on the LAMBDA integer transform. This paper draws on the idea of integer transform-down correlation in the LAMBDA method to do integer transform-down correlation on the float solution and variance matrix of the original ambiguity and positioning solution. Then the error in the ambiguity vector is identified and constrained based on the transformed variance matrix. If it can pass the test, the original integer float estimator calculation method is maintained, otherwise, the transformed ambiguity, positioning solution, and the corresponding variance matrices are filtered and brought into the original integer float estimator formula to calculate the positioning result. The method achieves the purpose of improving precision by filtering the coarseness and reducing the interference of coarseness on ambiguity. The proposed method is validated by the measured data, and the example results show that the method effectively weakens the error and improves the stability of the positioning results, and the spatial true error and RMSE of 3D positioning are improved by 16.1% and 14.7%, respectively, compared with the original integer float estimator, which ensures the robust application of the integer float estimator.