本文对基于最大后验概率密度函数（ｐ．ｄ．ｆ）的盲均衡技术作了研究，提出了一种实用的软判决引导的盲均衡算法，该算法能够得到快速的收敛速度和小的剩余码间串扰（ＩＳＩ）；并提出了使用软判引导算法的盲判决反馈均衡器（ＤＦＥ），其性能明显优于线性均衡器（ＬＥ）；最后，保证新算法可靠地收敛，给出了卷积噪声方差的近似迭代算法．模拟结果证明了提出的均衡技术对各种信道和ＱＡＭ信号的有效性． In this paper, the blind equalization technique based on maximum a posterior probability density function (p.d.f) is studied, and a practical soft decision-directed blind equalization algorithm is proposed, which can obtain fast convergence speed and small residual (ISI) is proposed. The blind decision feedback equalizer (DFE) using the soft decision guide algorithm is proposed, whose performance is obviously better than that of the linear equalizer (LE). Finally, the new algorithm is guaranteed to converge reliably. Approximate Iterative Algorithm of Product Noise Variance. Simulation results demonstrate the effectiveness of the proposed equalization techniques for various channels and QAM signals.