Hadi Zayyani - Department of Electrical and Computer Engineering Qom University of Technology Qom, Iran
Mehdi Bekrani - Department of Electrical and Computer Engineering Qom University of Technology Qom, Iran

This paper presents an improved continuous mixed p -norm (ICMPN) algorithm for sparse system identification under impulsive noise environment. The cost function of continuous mixed p -norm isregularized by a soft 0 l -norm term for increasing the convergence rate or equivalently decreasing the steady-state error. Simulation results show the effectiveness of the proposed algorithm in comparison to several of its well-known counterparts

Adaptive filter, System identification, Sparse system