Real Time Prediction of GPS Receivers Timing Errors using Parallel-Structure Neural Networks

  • M.R Mosavi Department of Electrical Engineering, Behshahr University of Science and Technology, Behshahr 48518-78413 Iran

要旨

GPS has become the primary system for distributing time and frequency. The signals are available nearly anywhere on earth, and provide a convenient link for establishing traceability to national and international standards. This paper presents two types Parallel-Structure Neural Networks (PSNNs) based on Wavelet Neural Networks (WNNs) and Recurrent Neural Networks (RNNs) for accurate modeling and prediction of GPS receivers timing errors. The PSNNs consist of a multiple number of component NNs connected in parallel. Each component NN in the PSNN predicts future data value independently based on its GPS receivers past timing errors data with different embedding dimension and time delay. An embedding dimension determines the number of inputs of each component NN and a time delay describes the interval of inputs of the GPS receivers timing errors. According to the embedding dimension and the time delay, the component NN takes various input-output pairs. The PSNN determines the final predicted value as an average of all the outputs of the component NN. We use actual data to evaluate the performance of the proposed methods. An experimental test setup is designed and implemented for this purpose. Results using the two methods are discussed. The experimental results obtained from a Coarse Acquisition (C/A)-code single-frequency GPS receiver strongly support the potential of the method using PSNNs based on WNNs to give high accurate timing. The GPS timing RMS error reduces to less than 46 and 177 nanoseconds, with and without SA, respectively.
出版済
2007-09-01
セクション
Article