Aiming at the problem of significant Doppler frequency shift (DFS) in the waveform of high altitude electromagnetic pulse (HEMP) emitted by aircraft moving at high speed, which severely reduces the accuracy of HEMP waveform parameter estimation, a method for perception-based DFS-assisted HEMP waveform parameter estimation was proposed based on the concept of integrated sensing and communication (ISAC). Firstly, the motion information of the target aircraft was perceived based on a multi-base radar system (MRS) to form the DFS of the HEMP waveform. Secondly, a Bayesian correction method was utilized to improve the perceived DFS. Finally, the HEMP data was compensated based on the corrected DFS to form a perception-Doppler-aided waveform parameter estimation method. The results show that the proposed method reduces the relative error in HEMP waveform parameter estimation and alleviates the impact of DFS on the estimation; at the same time, the proposed Bayesian correction method effectively enhances the accuracy of perceived DFS. Under different aircraft speeds and various multipath environments on the ground, the proposed method still improves the accuracy of HEMP waveform parameter estimation, which provides reference for seeking new methods for further improving HEMP waveform parameter-estimation accuracy.