Selective Availability via the Levinson Predictor

Timothy Barnes

Abstract:	The use of a generalized Levinson Predictor to provide estimates of future values of Selective Availability (SA) is explored.The prediction of SA provides the DGPS user a means for overcoming the latency of the corrections. The predicted estimates are derived from a linear combination of prior measurements of SA. An iterative method for deriving the weights, choosing the interval between data samples against which the weights are applied, and gener- ating an estimate of the mean square error of the predic- tor.The accuracy of the estimate is presented as a function of prediction interval. The RMS of the errors in predicting future values of SA using a Levinson predictor are approximately 0.2,2.0, and 4.0 meters for predictions of 8,30 and 60 seconds using code ranging data smoothed with carrier phase. The weights are determined from an estimate of the SA auto- correlation made at the beginning of each collection pass. The SA process tends to be wide-sense stationary in that the weights determined from one pass can be applied to future passes without significant degradation of prediction accuracy. Additionally, a set of weights determined from one PRN can be applied to other PBN’s without significant degradation in predictor accuracy. Finally, the user, with a modest amount of computation, can select the prediction interval desired given past SA values and an autocorrelation function.
Published in:	Proceedings of the 8th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1995) September 12 - 15, 1995 Palm Springs, CA
Pages:	533 - 542
Cite this article:	Barnes, Timothy, "Selective Availability via the Levinson Predictor," Proceedings of the 8th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1995), Palm Springs, CA, September 1995, pp. 533-542.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In