Future GNSS Acquisition Strategies and Algorithms

Abstract

Global Navigation Satellite Systems (GNSS) today face challenges such as signal path loss, interference, jamming, and timing inaccuracy in receivers. These challenges can be mitigated and GNSS systems can be modernized by the development of Low Earth Orbit (LEO) GNSS satellites that transmit signals on much higher frequencies and with much wider bandwidths. In this paper, we assess the feasibility of these changes from the point of view of signal acquisition. We investigate the challenges to acquisition that arise due to these changes, and we find that the most significant challenge is a dramatic and potentially prohibitive increase in acquisition time. We then attempt to use computational methods to reduce acquisition time. The Galileo E5 AltBOC(15,10) signal is used as a model wide bandwidth signal, and one satellite transmitting this signal is acquired from a live-sky observation data set using traditional acquisition techniques. To improve acquisition time, we implement a circular frequency shift algorithm, and we run the tested acquisition algorithms on a graphics processing unit (GPU). We attempt AltBOC signal acquisition with both techniques independently and together, and we find that when used together, acquisition time can be reduced by about 40%. Because acquisition time can be reduced, we conclude that acquisition is adaptable to the proposed changes to GNSS systems and thus, they are feasible from this point of view. We also attempt to create a computational complexity model to understand the complexity of acquisition in terms of computer operations and how time reduction techniques might also reduce computational complexity. We develop a model and find that it accurately represents the computational complexity and time of acquisition on a central processing unit (CPU) but does not accurately represent the computation time on a GPU, because it does not capture the high efficiency of the GPU.