Variable step-size modified blind equalization algorithm based on fractional lower order statistics under impulsive noise

In order to improve the performance of blind equalization under impulsive noise environments, a novel variable step-size modified blind equalization algorithm based on fractional lower order statistics is proposed. In the proposed method, fractional lower order statistic (FLOS) is used as cost function to update the weight coefficient of equalization, which can suppress the impulsive noise effectively. Moreover, in order to solve the problem of phase rotation, the cost function is modified by the imaginary part of FLOS, which can supplement the phase information. Furthermore, a variable step-size function is defined to improve the convergence speed. The results of Monte Carlo simulations show that the algorithm converges faster and suppresses impulse noise effectively at the same time.