在无线通信信道中,码间干扰严重影响了通信的质量和可靠性。传统的自适应均衡算法需要训练序列降低了传输效率,恒模算法(CMA)可以很好的克服了这个问题。针对CMA算法收敛速度较慢和无法纠正相位误差的缺点,提出一种新的步长迭代方法的改进型恒模算法(NSMCMA)。蒙特卡罗仿真表明,提出的算法,相比传统CMA算法,不仅可以纠正信道的相位偏转,而且具有更好的均衡效果和收敛性能。 In the wireless communication channel, intersymbol interference seriously affects the quality and reliability of communication. The traditional adaptive equalization algorithm needs training sequence to reduce the transmission efficiency. The constant modulus algorithm (CMA) can overcome this problem well. Aiming at the shortcoming of CMA algorithm converging slowly and unable to correct the phase error, a new improved constant modulus algorithm (NSMCMA) is proposed. Monte Carlo simulation shows that the proposed algorithm can not only correct the phase deflection of the channel, but also has better balance effect and convergence performance than the traditional CMA algorithm.