Cutoff rate of coded FH/MFSK with neural network jamming state information generator

A neural network approach to jamming state information (JSI) generation is presented. A coded frequency-hopped M-ary frequency-shift-keying system under partial-band noise jamming, plus additive white Gaussian noise is considered. Two cases of the neural network JSI generators are simulated: (1) with and (2) without knowledge of partial-band jamming fraction. Cutoff rate performance criteria are employed. The neural network JSI generators can offer significantly better performance than a JSI generator with a maximum a posteriori rule based on the total energy received in the M-tone signaling bands