正交小波分数间隔频率分集自优化盲均衡算法

A Self-optimized Blind Equalization Algorithm Based on Orthogonal Wavelet Transform and Fractionally Spaced Frequency Diversity

为克服传统多输入判决反馈盲均衡算法(MI-DFE)稳态误差大、收敛速度慢的缺点,在分析正交小波、分数间隔均衡器、频率分集技术的基础上,提出了一种正交小波分数间隔频率分集自优化盲均衡算法(WT-FF-SOC).该算法将正交小波、分数间隔均衡器与频率分集技术相结合,先得到正交小波分数间隔频率分集盲均衡算法(WT-FF);对WT-FF中每一路的输出合并后再进行判决反馈,用常数模算法(CMA)更新权向量,并作为WT-FF-SOC的跟踪模式;对WT-FF中的每一路信号进行判决反馈处理后,再对判决反馈滤波器的输出进行合并,在推导盲RLS算法后,用盲RLS算法更新权向量,并作为WT-FF-SOC的启动模式。两种模式利用代价函数的判决阈值进行切换。该算法具有收敛速度快、稳态误差小的特点。水声信道的仿真结果,验证了算法的有效性。

For overcoming the large mean square error （MSE） and slow convergence rate of traditional multiple-input decision-feedback blind equalization algorithm （MI-DFE） , a self-optimized blind equalization algorithm based on orthogonal wavelet transform and fractionally spaced frequency diversity （WT- FF-SOC） was proposed on the basis of analyzing orthogonal wavelet transform, fractionally spaced equalizer, and frequency diversity technique. In the proposed WT-FF-SOC algorithm, fractionally spaced frequency diversity blind equalization algorithm using orthogonal wavelet（WT-FF） was first obtained through fusing orthogonal wavelet,fractionally spaced equalizer, and frequency diversity technique; after the out- put of every branch in the WT-FF was firstly combined, the structure of decision feedback based on the combined signal was introduced, and the tracking mode of WT-FF-SOC with constant modulus algorithm （CMA） was established; and after the every branch in the WT-FF introduced decision-feedback filters, the outputs of the decision-feedback filters were combined, and this structure was used as the starting up mode of WT-FF-SOC with blind RLS algorithm. Two modes can automatically switch according to the threshold of cost function decision. The proposed WT-FF-SOC algorithm has better ability to improve the convergence rate and reduce the MSE. The simulated results in underwater acoustic channel show validity of WT-FF-SOC algorithm.