The present application relates to the field of communication, in particular to a method and device for double optimization of cognitive interference waveform and power based on
reinforcement learning, equipment and medium, according to the configurable parameters of the interference device, the interference waveform
library containing multiple interference waveforms is constructed; according to the current
electromagnetic environment state information of the interference target to be interfered by the interference device, the optimal interference waveform is determined from the interference waveform
library; under the premise that the optimal interference waveform remains unchanged, according to the current jam-to-
noise ratio state of the interference target, the lowest power meeting the effective interference threshold is determined; according to the optimal interference waveform and the lowest power, the interference
signal is generated. Through the technical means of constructing the interference waveform
library, determining the optimal interference waveform according to the
electromagnetic environment state, determining the lowest power meeting the effective interference threshold according to the jam-to-
noise ratio state, and generating the interference
signal according to the optimal waveform and the lowest power, the joint optimization of waveform and power is realized.