A robust constant modulus algorithm in alpha-stable noise environments

To improve the performance of the blind equalizer in alpha-stable noise environments, a robust constant modulus blind equalization algorithm is proposed. The error function of the original algorithm is improved and nonlinearly transformed, thus the affection of alpha-stable noise is effectively suppressed. Using shallow water acoustic channel, computer simulations were performed under Gaussian and alpha-stable noise respectively. The simulation results demonstrate that the new algorithm achieves the same convergence rate and lower residual intersymbol interference compared with the original algorithm in Gaussian noise environments. In alpha-stable noise environments, the original algorithm becomes unstable, while the proposed algorithm keeps robust and converges stably to the low intersymbol interference.