Abstract: Data reusing normalized least mean squares (DRNLMS) algorithms converge often faster than the conventional least mean squares (LMS) algorithm. This paper analyzes an adaptive DRNLMS, ADRNLMS, algorithm which has lower computational complexity relative to DRNLMS algorithm. Convergence behavior of an ADRNLMS algorithm are theoretically derived and analyzed. A large number of reusing times was found to raise the convergence rate but also increase computational complexity. In the proposed ADRNLMS algorithm is shown that number of reusing time is related to boundary of selected error. Decreasing of estimation error from selected threshold is caused decreasing of number of reusing time and vice versa. Simulation results validate the analysis and ensuing method.

Keywords