Read and translate Text c using Essential Vocabulary Text c. Adaptive Filter Operation

The primary function that is performed by the adaptive filter is system modeling. This is illustrated in Figure 2, where a spectrally white primary signal is supplied direct to either the s or y input of the adaptive filter. The primary signal is also input to a system with impulse response, H(n), and its output is connected to the other adaptive filter input. Two different options may be then implemented which result in quite distinct optimum weight vectors, H_opt, in the adaptive filter. These occur when:

The unknown system, H(n), is in the “y” input path to the adaptive filter /Figure 1 /. Here the optimum impulse response of the adaptive filter is a direct model of the system response, H(n).

Figure 2. Application of adaptive filter to (a) direct system modeling and (b) inverse system modeling.

Alternatively, the unknown system, H(n), can be inserted in the “s” input path to the adaptive filter /Figure 2(b)/. In this case the optimum response of the adaptive filter is the inverse of the unknown system response.

A practical example that illustrates adaptive filter operation in the first mode (i.e., direct system modeling) is in echo cancellation across a telephone line hybrid.

An example that may be used to illustrate operation of the adaptive filter in the inverse response modeling mode of operation is that of equalization for distortion in data transmission over telephone lines. Here, the input to the telephone line is excited by a known signal and the distorted line output is supplied to the s(n) input of the adaptive filter. The filter is then trained by supplying a replica of the known (perfect) primary signal at the y(n) input. The adaptive filter models the inverse of the line response to output equalized (distortion- free) data.

A further area where adaptive filters are applied is for noise cancellation /Sondhi and Berkley/. In this configuration the primary signal which contains the desired information contaminated by a spurious signal, is applied to the y(n) input.

A separate, correlated sample of this spurious signal is then obtained from another source which does not contain any of the required signal components. If this correlated signal is applied directly to the s(n) adaptive filter input, the filter forms an impulse response that yields an output y(n) which subtracts coherently the unwanted component from y(n) leaving only the desired signal on the e(n) output.

One example of the use of this technique is in the monitoring of fetal heartbeat /Widrow et al. 1975/. The primary signal is obtained from a transducer on the mother's abdomen. This transducer yields a signal containing the fetal heart signal but it is heavily masked by the mother's heartbeat. The secondary signal registering only the mother's heartbeat is then obtained from a second transducer on the mother's chest. The adaptive filter then models the distorting path from the chest to the abdominal transducer to produce the signal that coherently subtracts from the abdominal signal. Further examples are in the removal of engine noise from a pilot's microphone in an aircraft cockpit or analogous applications in other acoustically noisy environments such as in large power plants.

Another application of adaptive filters is in realizing a self-tuning filter which can be used to detect a sinusoid that is obscured by wideband noise. This application to adaptive line enhancement (ALE) /Zeidler et al./ is realized by feeding the signal directly into the y(n) input of the filter and a delayed version of the signal to the filter s(n) input. Provided that the delay is longer than the reciprocal of the filter bandwidth, the noise components will no longer be correlated at the two filter inputs but the sinusoidal components will still be correlated. The adaptive filter provides a sinusoid with an enhanced signal-to-noise ratio at its output, while the sinusoidal components are reduced at the error output.

Task 2. Complete the following sentences using the terms given below.

1. The primary function that is performed by the adaptive filter is ………………..

2. A spectrally white primary signal is supplied direct to …………….or…………...

3. The primary signal is also input to a system with…………………………………

4. When the unknown system, H(n), is in the “y” input path to the Adaptive Filter, the optimum impulse response of the Adaptive Filter is a ………………………………

5. When the unknown system, H(n), is inserted in the “s” input path, to the Adaptive Filter, the optimum impulse response of the Adaptive Filter is the ………………………………

6. A further area where Adaptive Filters are applied is for……………………………..

Terms to be inserted: inverse of the unknown system response; system modeling; noise cancellation; either the “s” or “y” input; impulse response H(n); direct model of the system response H(n).

Task 3. Look at Fig. 2 (page 14) and try to describe the application of adaptive filter in the models using the following terms and common words:

unknown system H(n); to be in / to be inserted in; “y” / ”s” input path; here / in this case; optimum impulse response; to be a direct model of the system response H(n); to be the inverse of the unknown system response.

Task 4.Answer the questions to Text C.

1. What is the primary function performed by the Adaptive Filter?

2 .What two different models of adaptive filter operation are applied?

3 .What practical examples that illustrate adaptive filter operation do you know?

4 .What is a further area where adaptive filters are applied?

5. How can you describe the scheme of noise cancellation?

Scheme A and common words can be used.

Scheme A. Noise cancellation.

* spurious – мешающий;

Common words to be used to describe the scheme:

1. to contain

2. to be applied

3. to produce

4. to be applied

5. to form

6. to substract the unwanted component

7. to leave the desired signal

Task 5. Put all possible questions to the following statements to discover all the details.

Liner distortions arise in many different ways in telephony networks. Three distinct types can be identified: amplitude distortion, group-delay distortion, and echoes.

Task 6. Translate the text into English.

Адаптивные фильтры применяются для устранения шумов двигателя в микрофоне пилота в кабине самолета.

Также адаптивные фильтры могут использоваться для подавления акустических шумов окружающей среды, например, на крупных электростанциях.

Еще одно применение адаптивных фильтров - это реализация самонастраивающегося фильтра. Этот фильтр используется для выделения синусоиды из широкополосного шума.

Task 7. Grammar Training. Translate the following sentences. Pay attention to the “Complex Object”.

1. We know the primary function of Adaptive Filters to be a system modeling.

2. They believe the primary signal to be input to a system with impulse response H (n) and its output to be connected to the other adaptive filter input.

3. He thinks a practical example of the 1-th mode to be echo cancellation across a telephone line hybrid.

4. The specialist assumed the adaptive filter to model the inverse of the line response to output equalized (distortion-free) data.

5. The students heard the adaptive filter be applied for noise cancellation.

6. Scientists know an example that may be used to illustrate operation of the adaptive filter in the inverse response modeling mode of operation to be equalization for distortion in data transmission over telephone lines.

7. They found the input to the telephone line to be excited by a known signal and the distorted line output of the adaptive filter.

8. We supposed them to use the adaptive filters for realizing a self-tuning filter which can be used to detect a sinusoid that is obscured by wideband noise.