A new learning algorithm based on MCC for colored noise interference cancellation

作者：Qu Hua^*; Ma Wentao; Zhao Jihong; Wang Tao

来源：Journal of Information and Computational Science, 2013, 10(7): 1895-1905.

DOI：10.12733/jics20101514

• Adaptive filtering algorithms
• Adaptive iterative learning
• Colored noise
• Constant values
• Correntropy
• Gradient descent algorithms
• Noise interference
• Step-size controls

In view of the defect of adaptive filtering algorithm based on the LMS in colored noise interference cancellation, we put forward a kind of new adaptive iterative learning algorithm using correntropy and Maximum Correntropy Criterion (MCC). The MCC and the gradient descent algorithm are used to update the weights of filter. The proposed algorithm can obtain an optimal factor of step length μ(k) from the square error of angle to undertake an analysis, and in order to illustrate the feasibility of the algorithm, the convergence of the algorithm from the sequence of the convergence in norm were analyzed. At the same time, a step-size control parameters was introduced in the algorithm, according to the parameters set constant values or variable values, the algorithm is defined as NC-MCC (Normalized Constant-MCC) and NV-MCC (Normalized Variable-MCC). The simulation results show that, in the signal adding the colored noise, compared with NLMS algorithm, the proposed algorithm is better and has good stability.

• 出版日期2013