1: \begin{abstract}

2: This letter proposes a novel adaptive reduced-rank filtering

3: scheme based on joint iterative optimization of adaptive filters.

4: The novel scheme consists of a joint iterative optimization of a

5: bank of full-rank adaptive filters that forms the projection

6: matrix and an adaptive reduced-rank filter that operates at the

7: output of the bank of filters. We describe minimum mean-squared

8: error (MMSE) expressions for the design of the projection matrix

9: and the reduced-rank filter and low-complexity normalized

10: least-mean squares (NLMS) adaptive algorithms for its efficient

11: implementation. Simulations for an interference suppression

12: application show that the proposed scheme outperforms in

13: convergence and tracking the state-of-the-art reduced-rank schemes

14: at significantly lower complexity.

15: 

16: \end{abstract}