Complete Software Simulation of the GPS Navigation Signal — Can It Be Good Enough to Fool Mother Nature?

Ted Driver

Abstract:	The GPS navigation signal, carrier, pseudo-random noise code (PRN) and navigation data, are inherently hardware- based. The carrier and PRN code are based upon the stability of an atomic clock, a very accurate and delicate piece of hardware. The stable frequency of the clock drives many hardware counters and registers, producing, with unerring accuracy, the PRN code we use to navigate. Is it possible then, for a software simulation of the GPS navigation signal to be good enough to fool ‘Mother Nature’, the Master Control Station (MCS)? This paper will investigate this premise, describing how a software simulation of the navigation signal can be constructed and the accuracy that must be met. To do this simulation justice, a full constellation of GPS satellites is needed, each sending their signals from their correct position in simulated space. A full array of monitor station models accurately spaced around the simulated earth are also needed, to supply the MCS with the required pseudorange (PR) and phase data. The MCS will judge, utilizing Kalman Filter smoothed PR measurement residuals, whether or not it is receiving data from its GPS constellation, or from a very good simulation. The simulation models consist of data from the navigation payload on board the satellite, environmental models to simulate signal delays, and noise models to make the simulation appear real. There are no radio frequency signals produced since there are no real receivers within the simulation. The navigation ranges are calculated for each satellite/ground location pair and then sent to the MCS in a format identical to that sent by the monitor stations. The MCS continuously receives PR and phase data from the simulated monitor stations, and processes the data every 15 minutes using a Kalman Filter, to predict new states for the constellation. The results of the simulation are unprecedented. The answer to the title question is yes, the MCS can be made to think it is processing real, hardware-generated data from a live GPS satellite constellation. After several 15 minute periods (known as K-points), most PR residuals are under 4 meters, leaving no doubt that a software simulation of GPS hardware signals is not only possible, but accurate.
Published in:	Proceedings of the 55th Annual Meeting of The Institute of Navigation (1999) June 27 - 30, 1999 Royal Sonesta Hotel Cambridge, MA
Pages:	49 - 55
Cite this article:	Driver, Ted, "Complete Software Simulation of the GPS Navigation Signal — Can It Be Good Enough to Fool Mother Nature?," Proceedings of the 55th Annual Meeting of The Institute of Navigation (1999), Cambridge, MA, June 1999, pp. 49-55.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In