Abstract: A carrier frequency estimation algorithm based on the expanded autocorrelation in additive white Gaussian noise was proposed, Through theoretical analysis, the algorithm fully excavates frequency information contained by the autocorrelation function and derives a new expanded autocorrelation and improves the frequency estimation formula based on the Yan algorithm. Compared with the existing algorithm, Numerical simulations indicate the estimation variance of the proposed algorithm is more close to the Cramer-Rao low bound（CRLB）when the sequence length is small or the SNR is low. and the calculation amount is smaller under the same estimate variance compared with the Yan algorithm when the sequence length is large or the SNR is sufficient.