Parameter Estimation And Applications For Wideband Underwater Acoustic Channels

Due to the requirement of economic and social development,as well as the purpose of national defense security,research on underwater acoustic science and technology has been paid close attention to by the countries all over the world since the 1900 s.Compared with the terrestrial wireless channels,the underwater acoustic channels has its particularity and complexity,which are mainly represented in the significant multipath effect,time-variant characteristic,strictly band-limited feature and so on.The underwater acoustic coherent communications have been developed consistently over the last two decades,since they provide greater spectral efficiency and higher data rate than the non-coherent communications.Meanwhile,the demand for channel estimation is increasingly high.From the signal processing perspective,the accuracy and efficiency of channel estimation have an immediate impact on the performance of underwater acoustic communication systems.Therefore,the channel estimation problem has always been one of the important priorities in the underwater acoustic research field.Starting from the physical characteristics of underwater acoustic channels,and seizing the wideband nature,this dissertation does deep research on the parameter estimation for multi-scale multi-lag underwater acoustic channels and its application on the communications.The main contributions are as follows.1)Combining the Ray theory and the analysis of underwater acoustic Doppler effect,this dissertation systematically introduces the wideband underwater acoustic channel model.For the channels which do not meet the narrowband condition,it is more effective to use the path-based Doppler scaling factors to describe the wideband Doppler effect.Thus the received signal can be represented by a superposition of multiple scaled and delayed versions of the transmitted signal,weighted by the wideband spreading function.Through some modification and improvement to the BELLHOP modeling tool and the Vir TEX algorithms,the multi-scale multi-lag underwater acoustic channels,which are under the cases of silent sea surface and sinusoid sea surface,can be simulated.2)Based on the existing wideband correlation processing methods for channel estimation,a kind of rational wavelet is adopted as the channel probe signal,then numerical algorithms of two channel estimation methods(namely the wideband deconvolution method and the matching pursuit method),are given.Simulation results show that both of the methods can provide good parameter estimates,when the wideband spreading function of the channel is sufficiently sparse.3)By utilizing the pulse compression of linear frequency modulation signals in both time domain and fractional Fourier transform domain,an accurate and efficient parameter estimation method,for the multi-scale multi-lag underwater acoustic channels,is proposed.It can separate the multipath components one by one according to the energy level.At each iteration,a sub-iteration is contained to adjust the optimal fractional order,so as to obtain more accurate estimates.Simulation results show that the estimation performance of wideband spreading function distinctly surpasses other existing methods.4)A spread spectrum communication scheme,based on the hyperbolic frequency modulation signals,is proposed for the multi-scale multi-lag underwater acoustic channels.In a communication frame,a linear frequency modulation signal is used as the frame head to estimate the channel parameters,and the data symbols are spread-spectrum modulated by a train of hyperbolic frequency modulation signals.The receiver has a very simple structure,and it can directly utilize the estimated channel parameters to deal with the correlation output according to independent multipath components.By specific diversity combination,the system is able to achieve the multipath-scale diversity gain.