A FUZZY SYSTEM FOR TONE DETECTION APPLICATIONS

In this paper we propose a computationally simple algorithm for the de tection of a sinusoidal tone for applications in the held of telecommu nication systems. More specifically, we have developed an algorithm ba sed on fuzzy logic (FL) to disable an echo canceller in a telephone ex change. The Fuzzy Tone Detector (FTD) presented uses three simple time -domain parameters and a matching phase based on only three fuzzy rule s, trained by a new hybrid learning tool. A fuzzy system called a Fuzz y Phase Reversal Detector (FPRD) was also trained for the detection of phase reversals. The new algorithm outperforms the traditional method s in terms of robustness to channel noise, working up to a signal to n oise ratio (SNR) of 0 dB. (C) 1998 Elsevier Science Inc. All rights re served.