Echo cancellation and applications - Essay Example

?Echo Cancellation and Applications:- Cancelling the effect of echo is called echo cancellation. Echo is the delayed and distorted voice of the original sound. There are certain techniques for the echo cancellation called the echo canceller, corresponding to the field of adaptive filtering. Their scope lies in the full duplex data transmission over wire to wire circuits. Echo is observed in telecommunications. Certain filter structures are used including FIR, IIR, lattice structure, frequency domain structure, memory type structure, among which FIR is of most use. Standard bodies such as CCITT evaluate methods and performance requirements for echo cancellers. Two algorithms such as Least squares method and Least Mean square (LMS) methods are used among both, the LMS is most widely used due to its characteristics. But its drawback is that it is correlation of the reference signal dependent. Echo is controlled from the delayed voices resulting from the hybrid transformers (two wire to four wire conversion) and from acoustics resulting from reflection or coupling of sound waves from loud speaker and microphone. The placement of the echo canceler largely depends on the frequency of the echo produced and the amount of echo cancelation required. For the telephone circuits, the FIR is widely used as an echo canceler, whereas for full duplex data models the echo canceler is placed where hybrid connects modem to two wire loop. In the acoustic echoes, signals are converted into many frequency bands due to long durations of the echo as FIR used as an echo canceler will require more complexities, whereas IIR and other cancellers are under study. For ISDN digital loop transmission, due to limited echo duration, either FIR is used or memory type echo cancelors are used. Hence we can conclude that the Echo cancellers provide an effective way for the echo cancelation and voice disturbances, and different type of echo cancelers are used in different cases and scenarios. The VLSI technology is opening new doors for the study. Whereas more study is needed regarding the acoustic echoes. Blind Signal separation using Steepest Descent Method:- Speech enhancers such as microphone arrays are used widely in the communication, consisting of two techniques, such as blind signal separation and beam forming technique. However there is a difference in two techniques because of which, the blind signal separation has a competitive advantage over the beam forming technique due to which, the blind signal separation is widely used compared to the beam forming technique. Because source localization and array geometry is required in the beam forming technique and unlike this technique, the sources are statistically independent. Two blind signal approaches are higher ordered based blind signal approaches and second order based blind signal approach. Due to the slow rate of convergence of the gradient based blind signal system, and for the convoluted mixture, a method is proposed that follows the steepest decent algorithm that improves the convergence characteristics. For optimum size, parabolic interpretation and modified Golden search procedure was observed. The Brent’s method was used for calculating the optimized step size. The Binary signal system is combined with adaptive noise canceler reducing further noise. Compared to the fixed step size gradient based algorithm, the binary signal system contributed to the higher convergence. The result also showed that the hybrid of binary signal system and the adaptive noise canceler provided higher noise suppressions. It was also concluded that the convergence rate was faster along with the desired cost function value. The experiments were performed in the simulated environments and the real room environments. Hence, we can conclude that the binary signal system is an effective technique to achieve the higher convergence. Whereas, a hybrid experiment also, was an effective effort when the binary signal system was combined with the adaptive noise cancelor. The complexity of the fixed step up sized could be improved by the decent adaptive optimization procedure. Acoustic Echo Canceler with High Speed Quality:- Due to the audio teleconferencing, the acoustic echoes are studied widely. The echoes are produced in the audio teleconferencing because of the coupling of sound waves between the microphone and the loud speaker. For hardware implementation different algorithms were studied such as learning identification (LI) and normalized least mean square (NLMS). A reductive correlation and variation loop gain were combined in this paper as a different algorithm. The echo cancelation characteristics of the learning identification and normalized mean square echo cancelations properties are challenged because of their efficiency. By combining the reductive correlation and the variable loop gain, a new algorithm was proposed. From the graphs, it was shown that the AP method has a better performance as compared to the LI method. For the convergence, the algorithm VA had the best results. The sub band echo cancelling was also performed by the technique by the signal division into sub bands. The simulation results were satisfactory with the analytical values obtained. Hence it is concluded that there are algorithms under study for the acoustic echo. Acoustic echoes produced in the audio teleconferencing are the subject of matter to be studied. As different algorithms are applied for this study, so there is still a need that the study is still under progress and a lot of remaining work has to be done regarding echo cancellation . And the new algorithm shows good results as compared to the other algorithms (conventional). The convergence characteristics of the new algorithms showed promising results. Furthermore, parallel echo cancellation architecture was also studied. Because of the presence of a module unit, this proposed architecture is implementable and is efficient in working. The NLMS and LI algorithms are also important for the study but the new developed algorithm showed more promising result as compared to the NLMS and LI algorithms. Acoustic Echo Cancellation Method Using Blind Source Separation Method:- In telecommunications, the acoustic cancellation is an important application that results from the coupling between the microphone and the loudspeaker. And also the algorithms play an important part to reduce this echo. Whereas the limitations of adaptive echo cancellation occur due to some other unwanted noise along with the echo. The blind source separation technique is an effective way for the echo cancellation. The echo signal and the received signal are separated, the step resulting in the increased the convergence and effective echo cancelation. The two microphones used for detecting echo and acoustic noise were assumed to be together, but separate them from each other and the echo signal was applied to reference signal. Acoustic noise was controlled by using blind source separation system. The results show the increased performance of the adaptive algorithm for the echo cancellation. Three computer simulation cases were considered. The first case included the white Gaussian noise was used as an echo signal and the speech signal as noise. In the next simulation, the white Gaussian noise was used as noise and echo signal as speech signal. And finally, the speech signal was used as an echo signal. The NLMS signal was used as an adaptive echo cancellation. Hence it was concluded that using a blind source separation system for echo cancellation is an effective approach. There is a poor performance because of the complexity formed when there are echo present along with the outer noise when mixed and simultaneously picked by microphone, the adaptive algorithm fails here to perform the task accordingly and shows poor conduct. The blind separation system is used and it is an effective way to separate the echo from the outer noise. Better echo cancellation is obtained by applying separated signal to the adaptive signal, hence resulting in better echo compensation. Acoustic Echo And Noise Cancelling Using Microphone Array:- The noisy car environment is an obstacle for the person in a car because the hands free performance is not up to the desired level. The microphone array is used in this case to increase the performance and decrease the unwanted echo observed in the car. The self calibrated microphone array would be a choice to follow in this situation. The onsite digital calibration data is a useful technique such as in this scenario of unwanted echo and noises. There are two phases in which the process is divided. The phases include the calibration phase and the adaptation phase. The microphones are controlled by LMS logarithm. The adaptive filters reduce the acoustic echoes and car noises coming from outsides. Another important task includes the speech quality inside the car. The array signals with signal to noise ratio is stored in a digital memory. The acoustic echo cancellation and the speech enhancement were planned to be made alongside. The microphone array system is effective in maintain the proper noise levels by controlling the loud speaker as well the car noises. Hence it is concluded that the adaptive array is useful for the echo cancellation and also along with it, there is also a trend observed of the noise suppression. The echo cancellation is an important aspect to be considered because of the unwanted noise produced. These echoes are not only to be removed but are a center of study because of health effects and safety effects mainly. The reason may be due to the echo distortion in car, there may be a possibility of accidents on the road, and this matter should also be studied along with the echo cancellation process Also, the near end speech has only slight distortion. And also the response of the array is not speed dependent nor to the construction of the road. Read More