Study On OFDM Channel Estimation Of Underwater Acoustic Communication System

Underwater acoustic (UWA) communication is a technology of sending and receivingmessages underwater, it been used in the military field originally. And it has been paid moreand more attention to while human beings has been incessantly exploiting and using themarine resources recently.Underwater acoustic channel bandwidth is limited，multipath interference is serious.Orthogonal Frequency Division Multiplexing (OFDM) technology is used to UWAcommunications; it can solve the multipath interference, enhance the utilization of bandwidth,and realize high data rate transmission. Meanwhile underwater acoustic channel has atime-varying and frequency selective decline characteristics, which in addition to need theappropriate design of OFDM systems, still must need channel estimation and equalization, sothat it can improve the performance of the UWA communication systems. Accurate channelestimation is the premise of the equalization.This paper mainly focuses on the UWA channel estimation and simple equalizationtechnologies in frequency domain of OFDM in UWA communications. The simpleequalization is to test and verify the accuracy of the channel estimation. Firstly, the UWAchannel characters, the realization form and basic model of OFDM are introduced, and thenthe reason of cyclic prefix (CP) combating multipath interference is analyzed. This papermainly discusses the channel estimation based on pilot, researches on block-type pilot patternfor slow fading channel and comb-type pilot pattern for fast fading channel according to thevariety rate of the channel, and simulates two pilot patterns combining UWA channelcondition. About block-type pilot pattern, investigate and simulate least square (LS), linearminimum mean square error (LMMSE) and Low-Rank-LMMSE algorithms. Aboutcomb-type pilot pattern, investigate and simulate linear interpolation, second orderinterpolation and discrete Fourier transform (DFT) interpolation based on improved LSalgorithm. About LMMSE, propose a simplified LMMSE algorithm. About LS algorithmdefects, obtain an improved LS algorithm based on noise reduction. At last, doexperimentations in water pail, LS and improved LS channel estimation algorithms areproved only.