A new approach to FM jammer suppression for digital communications

The paper considers FM jammer suppression in digital communications. It is pointed out that the cyclic autocorrelation function, commonly used for cyclostationary signals, and the ambiguity function, typically employed for analysis and recovery of nonstationary signals, such as FM, have the same formulation. However, nonstationary and cyclostationary signals have distinct localization properties in the time-lag frequency-lag domain, and this property can be effectively used for jammer suppression. Utilizing the spread of an FM jammer signature beyond that of the communication signal, one can select the jammer-only time-lag frequency-lag points for effective jammer spatial signature estimation. Suppression of the jammer signal is then proceeded by projecting the received signal to the jammers' orthogonal spatial subspace. Simulation examples for the recovery of BPSK signals in the presence of strong and moderate FM jammers are presented.