Acoustic feedback cancellation based on cepstral analisys

In a public address system, due to the amplification system, the source and loudspeaker signals are correlated. Acoustic feedback cancellation methods based on the Wiener theory present a bias in the estimate of the feedback path if no decorrelation algorithm is used. This work presents a new approach based on cepstral analysis to identify and cancel the feedback path. The only known previous cepstrum based formulations are taken in consideration and reviewed, and the process is now completely derived. A new method is proposed and simulation results showed that it is able to present an improvement in the estimation of the feedback path when compared to the state-of-art method.