Abstract: When the underwater acoustic channel is simultaneously affected by Doppler shift and multipath effects, the Doppler spreads of different propagation paths exhibit varying degrees of deviation; therefore, communication systems need to jointly estimate the delay and Doppler spread parameters of each path. However, existing methods face a difficult trade-off between computational complexity and estimation accuracy. Therefore, this paper proposes a channel parameter estimation method based on Doppler-invariant signals. Compared with hyperbolic frequency-modulated (HFM) signals, Doppler-invariant signals not only exhibit superior Doppler invariance but also possess a narrower main lobe and lower sidelobes in their autocorrelation functions. The algorithm proposed in this paper performs cross-correlation operations on two Doppler-invariant signals with opposite chirp directions in each iteration, followed by windowing and group-delay analysis to precisely align the correlation peaks—thereby yielding accurate estimates of the delay and Doppler spread for each propagation path. Numerical simulations and lake trials demonstrate that this method reduces computational complexity while significantly improving the accuracy of channel parameter estimation.