Analysis of application possibilities of Grey System Theory to detection of acoustic feedback

Notch-filter-based howling suppression is one of the most popular gain reduction method of dealing with acoustic feedback problem. The main goal of this paper is to analyze the possibilities of using the grey prediction model GM(1,1) in order to accelerate the feedback detection process of the algorithm. Computer based comparative simulations of the algorithm containing the prediction model in the detection stage and without it were performed. Simulations were performed for different prediction order, number of predicted samples and analysis window length. The comparison and evaluation were carried out for different source signals. Music, speech and noise signals were used.