Adaptive feedback cancellation in hearing aids with linear prediction of the desired signal

• Published in: IEEE transactions on signal processing 2005-10, Vol.53 (10), p.3749-3763
• Main Authors: Spriet, A., Proudler, I., Moonen, M., Wouters, J.
• Format: Article
• Language: English
• Subjects: Acoustic signal processing; Adaptation; Adaptive feedback cancellation; Algorithms; Applied sciences; Auditory system; Bias; Cancellation; Detection, estimation, filtering, equalization, prediction; direct closed-loop identification; Exact sciences and technology; Feedback; Hearing aids; identifiability conditions; Identification methods; Information, signal and communications theory; linear prediction; Loudspeakers; Mathematical models; Microphones; Neurofeedback; Predictive models; Signal and communications theory; Signal processing; Signal processing algorithms; Signal, noise; Studies; Telecommunications and information theory
• ISSN: 1053-587X; 1941-0476
• DOI: 10.1109/TSP.2005.855108

The standard continuous adaptation feedback cancellation algorithm for feedback suppression in hearing aids suffers from a large model error or bias if the received sound signal is spectrally colored. To reduce the bias in the feedback path estimate, we propose adaptive feedback cancellation techniques that are based on a closed-loop identification of the feedback path as well as the (auto-regressive) modeling of the desired signal. In general, both models are not simultaneously identifiable in the closed-loop system at hand. We show that-under certain conditions, e.g., if a delay is inserted in the forward path-identification of both models is indeed possible. Two classes of adaptive procedures for identifying the desired signal model and the feedback path are derived: a two-channel identification method as well as a prediction error method. In contrast to the two-channel identification method, the prediction error method allows use of different adaptation schemes for the feedback path and for the desired signal model and, hence, is found to be preferable for highly nonstationary sound signals. Simulation results demonstrate that the proposed techniques outperform the standard continuous adaptation algorithm if the conditions for identifiability are satisfied.