噪声鲁棒变步长LMS算法及其在OFDM水声信道均衡中的应用

针对最小均方(least mean square, LMS)算法在低信噪比(signal to noise ratio, SNR)条件下性能较差的问题,提出一种噪声鲁棒变步长LMS(noise robust variable step-size LMS, NRVSLMS)算法。该算法通过结合改进的双sigmoid函数和误差信号自相关函数,在迭代过程中动态调整步长的大小,解决了传统LMS算法中收敛速度、跟踪性能和稳态性能互相矛盾的问题。理论分析和仿真结果表明,与其他变步长算法相比, NRVSLMS算法抗噪声能力强,具有良好的跟踪速度和稳态性能。将该算法应用于正交频分复用(orthogonal frequency division multiplexing, OFDM)水声信道均衡中,与现有LMS类自适应均衡方法相比,基于NRVSLMS算法的信道均衡方法能够显著降低系统误码率(bit error rate, BER)和均方误差(mean square error, MSE)。

Noise robust variable step-size LMS algorithm and its application in OFDM underwater channel equalization

Aiming at the problem of poor performance in least mean square (LMS) algorithm under low signal to noise ratio (SNR) environments, a noise robust variable step-size LMS (NRVSLMS) algorithm is proposed. By combining the improved double sigmoid function and the error signal autocorrelation function, the NRVSLMS algorithm dynamically adjusts its step-size during the iteration process to solve the problem that the convergence speed, the tracking performance and the steady-state performance are contradictory in the traditional LMS algorithm. Theoretical analysis and simulation results show that the NRVSLMS algorithm has stronger anti-noise ability, faster tracking speed and better steady state performance compared with other variable step-size algorithms. The proposed algorithm is applied to orthogonal frequency division multiplexing (OFDM) underwater acoustic communications, the proposed equalization method base on the NRVSLMS algorithm improves the performance in terms of the bit error rate (BER) and mean square error (MSE) sufficiently in comparison with the existing ones.