Joint estimation algorithms based on LMS and RLS in the presence of impulsive noise

In order to solve the problem that the traditional adaptive algorithms are not able to perform well under the impulsive noise case,this paper develops new adaptive filtering algorithms in the presence of impulsive noise. A close inspection of the impulsive noise reveals that the noise has the sparse property in the time domain because it contains few large values and lots of small values in amplitude.By reformulating the cost functions utilizing this feature of noise into traditional adaptive algorithms,joint sparse online estimation algorithms are developed.The proposed algorithms exploit the noise structure to better suppress the noise.The results demonstrate the superior performance of the proposed methods compared to the existing p-norm algorithms in terms of convergence speed and steady-state error.