Low Computational Complexity in Low-Cost GNSS Receivers

One of the key issues in low-cost GNSS receivers is the computational complexity. One of the computational components of the GNSS receivers is the satellite positioning calculations. The main focus of this paper is to reduce the computational burden of this stage of processing. In this paper, four different models to fit the GPS and GLONASS satellites orbit are investigated. These models are compared with each other in terms of their computational load and accuracy, and the models have good accuracy and less computational load are selected. Among these four methods, the Hermite model and the Chebyshev model are superior to other methods for determining orbits of GPS and GLONASS satellites, respectively. In order to evaluate the performance of these models, two different data including ground stations data and measured data by GNSS receivers are used. The results show that these two models can improve the computational burden by about 90% compared to conventional methods like the Runge–Kutta and the Keplerian parameters that used in GNSS receivers.