Unsupervised Learning of GNSS Signals
Jing Ji, Wei Chen, Wuhan University of Technology, China; Hongyang Lu, Jiantong Zhang, Satellite Navigation Division of China Transport Telecommunication & Information Center, China
The ideal GNSS signal is structured. In this paper, we extracting, sampling and learning the GNSS raw signal, and optimizing the neural network through the basic functions. The prediction result of the learning network is compared with the output of the actual terminal at the same input terminal to verify the optimization effect. We believe that this approach allows the accuracy of GNSS (Global Navigation Satellite System) receivers to be close to the precision optimization bound in SDR (soft definition receivers) and is widely applicable to system optimization in other communications areas, further for PNT Information fusion.