www.ijraset.com
Vol. 1 Issue III, October 2013 ISSN: 2321-9653
I N T E R N A T I O N A L J O U R N A L F O R R E S E A R C H I N A P P L I E D S C I E N C E AN D E N G I N E E R I N G T E C H N O L O G Y (I J R A S E T)
Acoustic Echo Cancellation by Adaptive Combination of Normalized Sub band Adaptive Filters by Using Stochastic Gradient Algorithm Sunita, Parveen Bajaj Electronics & Communication Engg. Deptt. Maharishi Markandeshwar University, Sadopur Ambala, India
Abstract: Acoustic echo is a common occurrence in today’s telecommunication systems. It occurs when an audio source and sink operate in full duplex mode; an example of this is a hands-free loudspeaker telephone. In this situation the received signal is output through the telephone loudspeaker (audio source), this audio signal is then reverberated through the physical environment and picked up by the systems microphone (audio sink). The effect is the return to the distant user of time delayed and attenuated images of their original speech signal. The signal interference caused by acoustic echo is distracting to both users and causes a reduction in the quality of the communication. Adaptive filtering techniques are used to reduce this unwanted echo thus increasing communication quality. The input signal of the adaptive filter is highly correlated and the impulse response of the echo path is very long. These characteristics will slow down the convergence rate of the adaptive filter if the well-known normalized least mean-square (NLMS) algorithm is used. The normalized subband adaptive filter (NSAF) offers a good solution to this problem because of its decorrelating property, Which requires a tradeoff between fast convergence rate and small steady state mean-square error (MSE). The proposed combination is carried out in sub band domain and the mixing parameter that controls the combination is adapted by means of a stochastic gradient algorithm which employs the sum of squared sub band errors as the cost function. For the adaptation of the component filters, in addition to the conventional decoupling update method, we also propose a coupling one, which can further improve the performance of the adaptive combination scheme. Keywords: Acoustic echo canceller, Steady-state Mean-Square Error (MSE), Normalized least-Mean-Square (NLMS) algorithm, Sub band Adaptive Filters (SAFs), Normalized Sub band Adaptive Filters (NSAFs), hands-free telephone, teleconferencing system. 1.INTRODUCTION A common problem encountered in hands-free telephones and teleconferencing systems is the presence of echoes which are generated acoustically by the coupling between the loudspeaker and the microphone via the impulse response of a room. In recent years, there has been a great interest in the use of adaptive filters as acoustic echo cancellers to remove echoes. An adaptive filter can be characterized by its structure and adaptive filtering algorithm. The transversal filter with the well-known normalized least-mean-square algorithm is one of the most popular adaptive filters because of its simplicity and robust performance. In acoustic echo cancellation (AEC) applications, the speech input signal of the adaptive filter is highly correlated and the impulse response of the acoustic echo path is very long. These two characteristic will slow down the convergence rate of the acoustic echo canceller if the NLMS-based adaptive filter is used to remove echoes. One technique to solve the above problem is subband
Page 11
adaptive filtering. In conventional subband adaptive filters (SAFs), each subband uses an individual adaptive subfilter in its own adaptation loop, which decreases the convergence rate of SAFs because of the aliasing and bandedge effects. Lee and Gan presented a normalized SAF (NSAF) from the principle of minimum disturbance, with its complexity close to that of the NLMS-based adaptive filter. Its central idea is to use the subband signals, normalized by their respective subband input Variances, to update the tap weights of a fullband adaptive filter. Although these SAFs can obtain fast convergence rate when applied to AEC applications, they all require a tradeoff between fast convergence rate and small steady-state mean-square error (MSE) because of the use of a fixed step size. They all need to estimate the system noise power in advance. Recently, an adaptive combination of fullband adaptive filters has been proposed. The merit of this combination is that it can obtain both fast convergence rate and small steady-state MSE without estimate of the system noise