Research On Underwater Acoustic Channel Blind Equalization Algorithm Based On Orthogonal Wavelet Transform

In underwater communication systems, limited bandwidth and Multipath propagation can cause severe inter-symbol interference, Inter-symbol interference will cause transmission signal distorted, resulting in error at the receiving end, thus affecting the quality of communication.Conventional adaptive equalization need to send large amounts of training sequences frequently, seriously affecting the efficiency in the use of underwater acoustic channel. Blind equalization algorithm is using statistics of signal instead of sending a training sequence, thus suitable for the underwater acoustic channel.Thewavelet transform having better ability of distinguishing in time ang frequency domain provides a new approach for blind equalization algorithm. The signal correlation is effectively reduced through orthogonal wavelet transform, which can speed up the convergence rate of the equalization algorithm. The paper studies the theory of blind equalization for underwater acoustic channel blind equalization algorithm based on orthogonal wavelet transform algorithm, and verified by simulation the performance of blind equalization algorithm based on orthogonal wavelet transform.The main jobs of this thesis are follows:(1) The nature of orthogonal wavelet transform and orthogonal wavelet packet transform are studied. The equalizer output signal is expressed utilizing orthogonal wavelet and the orthogonal transform matrix to the equalizer input signal for different wavelet decomposition levels is worked out.(2) Firstly, the constant modulus algorithm based on orthogonal wavelet transform is studied, the algorithm performance of different wavelet and decomposition layers is analysised.Then, a variable modulus square contour algorithm based on orthogonal wavelet transform is proposed on the studing of the square contour algorithm, the new algorithm converges faster than the square contour algorithm,and inter-symbol interference is lower. To further reduce inter-symbol interference, Constellation match error is introduced into the variable modulus algorithm based on orthogonal wavelet packet transform; Finally, on the analysis of parallel soft decision blind equalization algorithm, Orthogonal wavelet based parallel soft decision blind equalization algorithm is proposed, the algorithm improves the convergence speed.(3) Firstly, the performance of constant modulus algorithm based on orthogonal wavelet packet transform is studied; Secondly, on the studying of high order QAM signal carrier recovery blind equalization, carrier recovery blind equalization algorithm based on the orthogonal wavelet packet transform is introduced to restore the phase rotation. Then, combined with the characteristics of the16QAM signal constellation, the paper proposes an improved blind equalization algorithm based on orthogonal wavelet packet transform, the algorithm not only reduces the inter-symbol interference but also has the capacity of correct phase rotation. Thirdly, in order to overcome the shortcomings of the fixed step, variable step size is introduced to the constant modulus algorithm based on orthogonal wavelet packet transform,Variable step size blind equalization algorithm based on the statistical measure and wavelet packet transform is proposed, Finally, an error function controlled variable step size blind equalization algorithm is proposed, the algorithm reduces the inter-symbol interference.