The Research Of Blind Adaptive Equalization Algorithm For Communication Systems

Blind adaptive equalizers play a key role in bandlimited digital communications systems in which the transmission of a training sequence is not possible or impractical. In high-speed wireless communication systems, equalization process is needed to suppress the intersymbol interference(ISI). Conventional equalization approaches use training signals to help channel equalization. When wireless channels, especially mobile channels, are not time invariant, training signals must be sent frequently. As such, much bandwidth has to be wasted by the training signals. Therefore, blind channel equalization methods have attracted much research interest because they can perform equalization without training signals, therefore achieving more efficient channel bandwidth usage.In this paper we focus on the constant modulus algorithm (CMA), which is a special Godard algorithm and, probably, the most popular blind equalization technique due to its simplicity. Then the orthogonal wavelet based CMA blind equalization (WBCMA) is analyzed. In comparison with CMA, the proposed algorithm shows a remarkable increase in convergence speed, but its cost is the increase in computational complexity. This paper applies quantification to WBCMA by using the integral power of 2 to quantify the error term. The bit of error term is decreased. Thereby, there is much reduction in computational complexity. Computer simulation proves the validity of this algorithm. A novel blind equalization algorithm based on stochastic gradient decent minimization of Renyi's entropy is introduced to orthonomal wavelet based CMA blind equalization (WBCMA). Compared with WBCMA, the proposed algorithms show a remarkable increase in convergence speed with only a moderate increase in computational cost. In this paper, we describe methods for computing fractionally spaced blind equalizers directly from second-order statistics of theobservations without channel identification, which reduce the performance degradation that is caused by channel estimation errors of channel estimation based approaches such as the subspace methods. On the other hand, compared with the CMA algorithm, ours use only second-order statistics; thus, faster convergence can be achieved.