Abstract: Unknown radar signal sorting is always a problem of radar countermeasure.So the paper proposes a new signal sorting method based on data potential field.The method can cluster signals based on the relevance of signal parameters and does not need to appoint cluster number in advance.This method can process mass of data without any prior knowledge and is not sensitive to the influence of noise.In the end computer simulation results show that the proposed approach can feasibly and effectively sort the radar emitter signals.

摘要： 未知雷达信号的分选一直是雷达对抗情报处理中的难题。因此，提出了一种新的基于数据势场聚类的未知雷达信号分选方法。该方法无需预先指定分类数，利用信号参数之间的关联性自动进行聚类，以此达到分选的效果。同时方法适宜于处理大量数据，对噪声影响不敏感，无需先验知识支撑。最后，通过计算机仿真验证了该方法的可行性和有效性。