Efficient Fast Stereo Acoustic Echo Cancellation Based on Pairwise Optimal Weight Realization Technique
- PDF / 1,793,400 Bytes
- 15 Pages / 600.03 x 792 pts Page_size
- 71 Downloads / 194 Views
Efficient Fast Stereo Acoustic Echo Cancellation Based on Pairwise Optimal Weight Realization Technique Masahiro Yukawa, Noriaki Murakoshi, and Isao Yamada Department of Communications and Integrated Systems, Graduate School of Science and Engineering, Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-Ku, Tokyo 152-8550, Japan Received 1 February 2005; Revised 1 October 2005; Accepted 4 October 2005 In stereophonic acoustic echo cancellation (SAEC) problem, fast and accurate tracking of echo path is strongly required for stable echo cancellation. In this paper, we propose a class of efficient fast SAEC schemes with linear computational complexity (with respect to filter length). The proposed schemes are based on pairwise optimal weight realization (POWER) technique, thus realizing a “best” strategy (in the sense of pairwise and worst-case optimization) to use multiple-state information obtained by preprocessing. Numerical examples demonstrate that the proposed schemes significantly improve the convergence behavior compared with conventional methods in terms of system mismatch as well as echo return loss enhancement (ERLE). Copyright © 2006 Masahiro Yukawa et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1.
INTRODUCTION
The ultimate goal of this paper is to develop an efficient adaptive filtering scheme, with linear computational complexity, to stably cancel acoustic coupling, from loudspeakers to microphones, occurring in telecommunications with stereophonic audio systems. This acoustic coupling is commonly called acoustic echo (we just call it echo in the following). The stereophonic acoustic echo cancellation (SAEC) problem has become a central issue when we design high-quality, handsfree, and full-duplex systems (e.g., advanced teleconferencing, etc.) [1–13]. A direct application of a monaural echo cancelling algorithm to SAEC usually results in unacceptably slow convergence [1–3], and this phenomenon is mathematically clarified in [5], showing that the normal equation to be solved for minimization of residual echo is often illconditioned or has infinitely many solutions due to inherent dependency caused by highly cross-correlated stereo input signals (see Section 2.2). Decorrelation of the inputs is a pathway to fast and accurate tracking of echo paths (impulse responses), which is necessary for stable echo cancellation [6, 8, 14, 15]. A great deal of effort has been devoted to devise preprocessing of the inputs [3, 5, 14–22] (see Appendix A). In other words, these preprocessing techniques relax the ill-conditioned situation with use of additional information provided artificially by feeding less cross-correlated input signals. Based on the
preprocessing [5], real-time SAEC systems have been effectively implemented, for example, in [8, 13]. Under rapidly time-varying situations, however, further convergence acceleration is strongly required
Data Loading...
