In order to estimate the time difference of arrival (TDOA) of frequency hopping signal accurately

the ambiguous characteristic of the correlation function is analyzed

and the TDOA estimation method of frequency hopping signal based on multi-hop coherent integration is investigated.The periodic peaks in the multi-hop correlation function will lead to ambiguous estimation of the TDOA;so two approaches to eliminate ambiguity are proposed:the approach of ambiguity elimination by TDOA averaging of single-hop (TASH) and by envelope fitting of correlation function (EFCF).Ambiguity elimination is achieved through fine peak searching between a small interval nearby the main peak

and the small interval is determined by the coarse TDOA obtained by means of TASH or EFCF.The difference between TASH and EFCF is that the first approach obtains the coarse TDOA by incoherent averaging of TDOA estimation of single-hop

while the second approach obtains the coarse TDOA by the envelope of correlation function which is fitted by using multiple periodic peaks found by whole searching.Numerical simulations demonstrate that the performance of the two proposed methods is obviously superior to the conventional CAF

and the requirement for the correlation output SNR of unambiguous estimation can be reduced

the frequency interval for unambiguous estimation can be increased.The EFCF has better performance but greater computational complexity.