In this paper

a nonlinear soft-thresholding method based on orthogonal bases of compactly supported wavelets in frequency domain is proposed to recover signal from transient scattered response contaminated by Gaussian white noise. Two poverties of this kind of wavelet basis are responsible for the effectiveness of the proposed method:one is its compact support in frequency domain

the other is its great capability in data compression. The effectiveness of the proposed method is verified with a synthetic transient scattered response

measured response of a sphere and response of a model aircraft B-58 having the scale ratio of 1: 72 with additive Gaussian noise. The results obtained by the method are compared with those of Fourier truncation

spline smoothing

maximum modulus algorithm of wavelet and nonlinear soft-thresholding based on compactly supported wavelet bases in time domain. The superiority of the proposed method is thus demonstrated.