In an aging society, the incidence of heart disease continues to increase, and among them, arrhythmia is a serious disease that accounts for 90% of sudden deaths and 20 to 30% of strokes, but the incidence is increasing. In this dissertation, we intend to deal with a non-contact, automated diagnosis method that uses UWB radar to diagnose arrhythmias more easily and conveniently and solves the shortcomings of the existing ECG method.