Abstract: A novel efficient algorithm for time difference of arrival (TDOA) and frequency difference of arrival (FDOA) estimation between two linear frequency modulated (LFM) signals is proposed in this paper. The proposed approach combines decimated ambiguity function and Radon transform to estimate the chirp rate of the LFM signal; then fractional Fourier transform (FrFT) is used to estimate the point of intersection between the ridge and Doppler axis; finally the peak of ambiguity function is searched efficiently along the ridge by using of dechirping. For the multi-component LFM signal, since different LFM signals have different ridges, the proposed approach can successfully distinguish each LFM signal from the received signal, and both the TDOA and FDOA of each LFM signal can be estimated precisely. Due to fast Fourier transform-based processing and the use of only one-dimensional searches, the proposed approach is computationally efficient. Simulation results show that with the increase of signal noise ratio (SNR), the variances of the estimates are gradually close to the Cramer-Rao lower bound.