Signal Sorting Algorithm of Hybrid Frequency Hopping Network Station Based on Neural Network

In non-cooperative frequency hopping communication system, the network station sorting of received hybrid signals plays an important role and becomes active research area in recent years. order to solve problem that currently widely used clustering algorithm cannot achieve satisfactory accuracy. this paper, we propose a signal method for stations by applying neural classify description words signals. Additionally, conjugate gradient is utilized training process improve convergence speed. Once finished, only one word input required obtain its own label real time. Simulation results demonstrate when compared with algorithm, proposed converges less iterations delivers better accuracy, especially low noise ratio environment.