Improve the position accuracy on low cost GPS receiver with adaptive neural networks

We study a way of using a low cost GPS receiver for position determination and propose a neural network for better positioning accuracy. First we define the GPS system errors. Then measuring the components of the position errors, a real and dynamic pattern of the errors is created and feed into the neural networks. These neural networks are taught with such real data to predict the errors of later seconds. The stages of neural networks implementation and the result of the tests are stated with real data. They show the errors of the position components decrease due to the training of the neural networks.