Adaptive equalization for underwater data transmission
暂无分享,去创建一个
The tracking behaviour of the fast RLS (recursive least squares) and LMS (least mean squares) algorithms is investigated in the multipath underwater transmission case. The convergence properties of the algorithms are discussed, and the algorithms are applied to an underwater experiment in a constant-time context. Several cases of fluctuating underwater data transmission are simulated by varying the typical multipath parameters: magnitude ratio of the paths, delays, phases and SNR (signal/noise ratio). The simulation results clearly indicate that the SNR is of a great importance in the tracking capability of an equalizer. For a low SNR, (5 dB, for example), an optimal FLS (fast least squares) algorithm is useless for tracking magnitude ratio or path phase variations. For a high SNR (>20 dB), a FLS algorithm can perform better for compensating fast magnitude variations. In the phase fluctuation case, the algorithms performed equally well.<<ETX>>
[1] T. Kailath,et al. Fast, recursive-least-squares transversal filters for adaptive filtering , 1984 .
[2] André Gilloire,et al. Methode de stabilisation des algorithmes des moindres carres rapides appliques au traitement de parole , 1987 .
[3] George Carayannis,et al. A fast sequential algorithm for least-squares filtering and prediction , 1983 .