Channel Estimation Based On Filtered Multitone For Underwater Acoustic Channels

Being considered as the primary long-distance wireless communication method in un-derwater situations,underwater acoustic communication plays an important role in various application areas.However,the study in underwater acoustic communication is still faced with problems.The propagation of acoustic waves is restricted by three main factors:the attenuation increased with signal frequency,the time-varying multipath characteristics and the relatively low propagating speed.Considering the above-mentioned problems together with the characteristics of multi-carrier modulation techniques,a multi-carrier modulation based on filter-banks,also known as Filtered-Multitone Modulation(FMT),is proposed in this paper.After taking the characteristics of underwater acoustic channels into consid-eration,the channel estimation and equalization technique based on FMT modulation is studied and proposed.In the following section,the feasibility of applying FMT modulation in underwater acoustic communication is firstly analyzed.Simulation result reveals its advantage over Orthogonal Frequency Division Multiplexing(OFDM)in combating the Doppler Effect.In FMT modulation,carriers are not strictly orthogonal to each other.The gaps between subcarriers lower the chance of Inter-Carrier Interference(ICI)and enhance the perfor-mance of FMT modulation when facing Doppler Effect.Secondly,the delay Doppler doubly spread characteristics of underwater acoustic communication channel are introduced,and a channel model based on the sparse charac-teristics of underwater acoustic channel is established.Several approximations have been made to make the model simpler and realizable.The actual channel is considered to be composed of several major propagation paths,and the time-varying amplitude and delay characteristics of the channel are approximated by polynomial fitting.Parametric modeling and channel estimation based on FMT modulation are carried out based on the model.By analyzing the input-output relationship of the parametric channel model,the high dimensional channel estimation problem can be transformed into the finite-scale param-eter estimation.Compression perception algorithm is adopted to simplify the calculation.Aimed at the time-varying amplitude and delay of the model,a step-by-step estimation method is proposed.In such method,the current channel characteristic is estimated using pilots at the beginning and end of the data sequence.Combined with the channel model characteristics,the model parameters can be obtained.Simulations are carried out to verify the performance of the method.Inter-Symbol Interference(ISI)caused by the delay spread in underwater acoustic communication is a major factor that limits the performance of shallow water acoustic communication systems.In FMT modulation,the effect of ISI cannot be ignored.To solve the problem,equalization techniques are usually applied at the receiver.In this paper,an adaptive Decision Feedback Equalization(DFE)based on Recursive Least Squares(RLS)algorithm is introduced.A Fractionally Spaced Equalizer(FSE)and a Doppler compen-sator based on Phase Locked Loop(PLL)are also employed to optimize the performance of the equalizer.In addition,a novel equalization method based on the ignorable ICI in FMT modulation is also proposed,in which the symbols on each subcarrier are equalized independently.The performance of both methods is verified through simulations.Finally,the reliability of FMT modulation is verified by sea trial data.