Abstract: It is significant for engineering application to study the fast frequency estimation algorithm.Thus,by establishing the mathematical models,an improved fast frequency estimation algorithm with short sequences is proposed for underwater acoustic signal detection at low signal-to-noise ratio(SNR).It is based on the three-point self-correlation algorithm,but provides more precision than the latter with no cost of increase in computational complexity.The simulation results show that the root mean square error(RMSE) performance of the proposed algorithm is closer to Cramer-Rao bound(CRB),which outperforms its counterpart by 5 dB at low SNR and reaches the CRB at the SNR up to 10 dB.