1: \begin{abstract}

2: %\boldmath

3: In order to improve the performance of Least Mean Square (LMS) based

4: system identification of sparse systems, a new adaptive algorithm is

5: proposed which utilizes the sparsity property of such systems. A

6: general approximating approach on $l_0$ norm -- a typical metric  of

7: system sparsity, is proposed and integrated into the cost function

8: of the LMS algorithm. This integration is equivalent to add a zero

9: attractor in the iterations, by which the convergence rate of small

10: coefficients, that dominate the sparse system, can be effectively

11: improved. Moreover, using partial updating method, the computational

12: complexity is reduced. The simulations demonstrate that the proposed

13: algorithm can effectively improve the performance of LMS-based

14: identification algorithms on sparse system.

15: 

16: \textbf{Keywords:} adaptive filter, sparsity, $l_0$ norm, Least Mean Square (LMS).

17: \end{abstract}