Combined Blind Equalization Algorithms Based On Wavelet Transform

In underwater acoustic communication system, the Inter-Symbol Interference (ISI), caused by multi-path effects and bandwidth limited channels, affect the performance of communication seriously. Unlike traditional trained equalization algorithm, the blind equalization operating without the aid of a training sequence can serve the bandwidth and improve the validity and reliability of communication, so it is very suitable for underwater acoustic channel equalization. Along with the wavelet transform theory speedy developing, a new transform domain blind equalization algorithm method appeared. The convergence rate can be speeded up via using the fast de-correlation ability of orthogonal wavelet transform. Aiming at the pronblems of various blind equalization algorithms, using wavelet transform as analyzing tool, blind equalization algorithms based on wavelet transform is deep studied in this dissertation. The main contributions are as follows:1 Analyzing blind equalization algorithms based on orthogonal wavelet transformAfter analyzing the factors affected the performance of blind equalization,we knowed that the equalizer input signal's autocorrelation is one of the important factors which affect the convergence rate. The autocorrelation of signal various components can be effictively reduced after carried on Orthogonal Wavelet Transform (WT) and normalizated the energy in wavelet transform domian. Hence,Fractional Spaced Equalizer Constant Modulus Algorithm based on WT (WT-FSE-CMA) and Super- Exponential Iterative Blind Equalization Based on Orthogonal Wavelet Transform (WT-SEI) are proposed, via analyzing the bases of the performance of WT-CMA.2 Analyzing combined blind equalization algorithms based on orthogonal wavelet transformFor making full using the advantage of each blind algorithm and eliminating the disadvantage,two classes wavelet transform domian combination algorithms has been researched.(1) Aiming at the character of the convergence of the liner equalization (LE) algorithm based on equalization criterion of statistics is fast,but its Mean Square Error (MSE) is larger. Using the soft-decision directed,we combined LE algorithm and Decision Directed (DD) by means of the decision to control the weigh vector of each part. Then,a novel combined equalization including the LE algorithm based on orthogonal wavelet transform (WT-LE) and DD (WT-LE+DD) is proposed.(2) The MSE of the Super-exponential Iteration (SEI) blind equalization algorithm is large and it can't correct phase rotation.A new combined Super-exponential Iteration blind equalization algorithm based on WT (WT-CSEI) is proposed. Towards the time-varing Doppler underwater acoustic channels,the proposed algorithm use the orthogonal wavelet transform to revise the weight vector iterative formula,and then introduce the phase-locked loop to it,combined with DD by using the soft-decision directed.3 Analyzing blind equalization algorithms based on orthogonal wavelet packet transformComparing with orthogonal wavelet transform,the autocorrelation of input signal is futher reduced via the signal scale space and wavelet space are decomposed respectively by Orthogonal wavelet packet transform. Hence,we integrated WPT with blind equalization algorithm,the CMA based on Orthogonal Wavelet Packet Transform (WPT-CMA) and SEI based on Orthogonal Wavelet Packet Transform (WPT-SEI) are proposed. Aiming at the distortion of nonlinear channels, a Decision Feedback Blind Equalization Algorithm Based on WPT (WPT-DFE) and its imporved algprithm are proposed.The efficiency of the proposed algorithms are proved by computer simulation in underwater acoustic channels.