Abstract: The spectra of acoustic Doppler Current profiler (ADCP) signal which is a non-stationary underwater sound reverberation process are estimated by adaptive lattice filter. Three programs, i. e.,stochastic gradient (SG), fast Kalman (FK) and recursive least square (RLS), are used. Learning curves are given in the case of optimum parameters. From these curves we can conclude that, RLS program is the best one. Its convergence speed is comparable with the one of FK, but its mean square error (MSE) is smaller than that of FK and SG. The spectra of ADCP signal are estimated with RLS program. The number of power spectrum Peaks may be one, three or two with increasing non-stationarity of medium movement,wthch correspond to the quasi-stationary area, non-stationary area and strong non-stationary area of medium movement, resistively. They are consistent with the theory of evolutionary spectra and Wigner-Ville spectra of moving medium.
In these theories, both the non-stationary factor A and characteristic width B are significant. When the time interval T corresponding to the volume element of the scatters is smaller than B, we can suppose that the power spectrum displays a single Peak characteristic, when T is in the neighbour of B, there are three Peaks in power spectrum, and when T is latgen than B, we may use the power spectrum which displays atwo Peak characteristics.