A Novel FSK Demodulation Method Using Linear Neural Network

摘要

A novel frequency shift keying (FSK) demodulation method using the linear neural network (LNN) is proposed in this paper. Even though the probability distribution function of the channel noise is unknown, the structure of the LNN based FSK demodulator can be derived. The design of the LNN and the demodulation scheme are dressed in detail. Theoretical analysis and computer simulation results show that the bit error rate (BER) performance of the demodulator is superior to that of the conventional coherent or incoherent FSK demodulator while the FSK signals are transmitted over an additive white Gaussian noise (AWGN) channel.