Session
Technical Session XI: Advanced Subsystems and Components II
Abstract
In this paper it is described the development of a receiver for GMSK signals based on the use of neural networks. Two cases are described: one in which the linear approximation of the GMSK signal is employed and other where phase changes are detected. Wide implementation, training and result differences are between both methods and they are described in the paper. The first approach is simpler than the second one, has an easier training but worst results from the point of view of error probability.
CPM Neural Network Based Receiver for LEO Satellites
In this paper it is described the development of a receiver for GMSK signals based on the use of neural networks. Two cases are described: one in which the linear approximation of the GMSK signal is employed and other where phase changes are detected. Wide implementation, training and result differences are between both methods and they are described in the paper. The first approach is simpler than the second one, has an easier training but worst results from the point of view of error probability.
