Focused on the problem that the precision of radar altimeter in low-altitude aircraft is affected by clutter interference in flight, the Karhunen-Loeve transform was proposed to apply in the anti-interference of radar altimeter. By utilizing different second-order statistic characteristics on the clutter and target signal in echo data, the target signal was successfully separated from the clutter according to the corresponding relation between the characteristic value of the autocorrelation matrix and each component in the echo data, and the accuracy of the altitude was improved. The simulation results based on ADS demonstrate the validity and practicability of the proposed method.