KR100640895B1 - Apparatus for fruquency domain equalizer using channel detector and Method for the same - Google Patents

Abstract

The present invention relates to a frequency domain equalizer using a channel detector. According to an aspect of the present invention, there is provided a frequency-domain equalizer of a superposition-storage scheme that overlaps and equalizes input data and extracts valid data from the equalized data, comprising: a channel estimator for estimating a channel impulse response of the input data; Provides a frequency domain equalizer using a channel detector configured to receive a channel impulse response to determine a data extraction position, and a block extractor for extracting valid data from the superimposed equalized data according to the determined extraction position. do. Therefore, by applying different positions for extracting valid data according to channel conditions, it is possible to increase the area for equalization without significantly increasing the hardware area.

Frequency domain equalizer, channel estimator, channel detector, block extractor

Description

Apparatus for fruquency domain equalizer using channel detector and Method for the same}

1 is a block diagram showing a configuration of a frequency-domain equalizer of a superposition-storage scheme according to the prior art.

2 illustrates an example of a channel response according to a general time delay.

3 shows the equalizer coefficient shape in a typical multiple channel environment.

Figure 4 shows the error at the last stage of a frequency domain equalizer in a typical multiple channel environment.

5 is a block diagram showing the configuration of a frequency-domain equalizer of the superposition-storage scheme using a channel detector according to the present invention.

6 is a block diagram showing an internal configuration of a channel detector according to an embodiment of the present invention.

7 is a block diagram showing an internal configuration of a channel detector according to another embodiment of the present invention.

8 is a diagram illustrating a frequency spectrum of a general channel.

-Explanation of symbols for the main parts of the drawing-

100: overlapper 200: FFT unit

300: channel information provider 400: complex multiplier

500: IFFT unit 600: block extractor

700: channel estimator 800: channel detector

The present invention relates to a channel equalizer, and more particularly, to a channel equalizer and a method for equalizing efficiently in the frequency domain using a channel detector.

In general, a digital signal transmission system is operated such that a signal sent from a transmitter is received at a receiver through a specific channel. In this case, when the channel equalizer is used, signals can be correctly received by eliminating echoes that may occur while passing through the channel.

In particular, in the case of terrestrial broadcasting, echoes that may occur during the signal transmission increase, and thus, when the channel characteristic has a long impulse response, the length of a filter used in the equalizer is also increased.

Increasing the length of the filter used in the channel equalizer is a factor that seriously increases the complexity of the hardware. In this case, an Infinite-duration Impulse Response (IIR) filter is generally used to equalize a channel having a longer response by using a filter of the same length, but the IIR filter is compared with a finite-duration impulse response (FIR) filter. There is a problem of poor stability.

Therefore, in order to stably equalize the response of a long and complicated channel, a long enough FIR filter is used, and the complexity of the required hardware is considerably increased.

The role of the filter is to perform a highly complicated convolution operation between the data input in the time domain and the coefficient of the filter. The complexity of the operation increases as the length of the filter increases.

However, this can be done simply by converting it to a simple multiplication operation if the data and coefficients are transformed into the frequency domain. This property enables channel equalizers to be implemented in hardware in the frequency domain.

In order to filter the signal in the frequency domain, the data is superimposed in the time domain and transformed into the frequency domain and then multiplied by an equalization coefficient, and the result is converted back into the time domain and the overlapped data is removed. Or a process of restoring the original signal by extracting overlapping amounts of data. This will be described with reference to the accompanying drawings.

1 is a block diagram showing a configuration of a frequency-domain equalizer of a superposition-storage scheme according to the prior art.

As shown in FIG. 1, a frequency domain equalizer according to the prior art generally uses an overlap-save scheme. This is a method of overlapping and calculating data to remove noise caused by the influence of a channel, and then removing equalized data to obtain equalized data.

To this end, when the input data u (n) enters the overlapper 10, a data block consisting of new data currently input and previously input data is constructed. In this case, the size of the data block is equal to the size of the fast fourier transform (FFT) unit 20 of the next stage.

As such, in the case of configuring a data block in which the old data and the new data have the same amount, the output speed of the data is twice the input speed. The length of the filter can be effectively changed by differentiating the ratio of old data and new data constituting one block of data, and the output speed of the overlapper increases in proportion to the amount of overlapping data. The superimposed data u (n) is converted into data U (k) in the frequency domain by the FFT unit 20.

Meanwhile, the channel information provider 30 obtains and outputs the equalizer coefficient W (k) of the frequency domain containing the channel information, and the equalizer coefficient W (k) is the data U of the frequency domain. (k) is multiplied by complex multiplier 40 to obtain equalized data Y (k). In this case, the equalizer coefficient W (k) in the frequency domain may be obtained by using channel information, or may be adapted to minimize an error in the equalizer output.

The data Y (k) of the equalized frequency domain is transformed into an inverse fast fourier transform (IFFT) unit 50 in the time domain, and the storage unit 60 subtracts the data corresponding to the overlapped amount of data in block units. Take the remainder to get the data y (n) of the final equalized time domain.

That is, the storage unit 60 performs the task of extracting valid data at a predetermined position of the calculated block and discarding the rest. The suitable position for extracting valid data is different depending on the characteristics of the channel. In general, as shown in FIG. 1, when the size of a block used in the equalizer is twice the size of the input / output data, a method in which the first half of the calculated block is discarded and the second half is taken.

This will be described with reference to the accompanying drawings, because it is assumed that the channel characteristics are generally the same as those of FIG. 2A when the valid data is taken in the same manner as the storage unit 60 of FIG.

In order to equalize the signal passing through the channel as shown in FIG. 2 (a), the coefficient of the FIR filter as shown in FIG. 3 (a) is required, and according to the result of one block by implementing the equalizer in the frequency domain It is shown as (a) of FIG. That is, as shown in (a) of FIG. 4, noise or errors exist in the front of the block, so that if the front part is discarded and the back part is taken, clean data can be obtained.

For reference, FIG. 2 shows an example of a channel response according to a time delay when the x axis is a time axis, FIG. 3 shows an equalizer coefficient in various channel environments, and FIG. 4 shows a final stage of a frequency domain equalizer of various channel environments. It is a figure which shows the error.

However, in the case of terrestrial broadcast channels, particularly in the case of receiving broadcasts in urban areas, the channels in the cases as shown in FIGS. 2B and 2C are also observed. In this case, the shape of the equalizer coefficients of the time domain required in each case is as shown in FIGS. 3 (b) and 3 (c), and the error included in the data block before using the equalizer in the frequency domain and removing the overlapped data is shown. The size of is generated like the shade of Figs. 4 (b) and (c).

According to the conventional method of extracting data from a fixed position, to solve this problem, the amount of overlapping data must be increased to increase the equalization area, which is calculated when calculating the same block size. This has to be reduced, resulting in increased power consumption and increased hardware speed. In addition, when the size of the processing block is increased to increase the overlapping data, there is a problem that the hardware size increases.

The present invention is to solve the above problems, by determining the state of the channel and determine the location of the data extracted accordingly, by increasing the area where the filter can be equalized without increasing hardware size and power consumption, It is to provide a frequency domain equalizer using a detector.

In order to achieve the above object, the present invention is to estimate the channel impulse response of the input data in the overlap-store frequency domain equalizer for overlapping and equalizing the input data and extracting valid data from the equalized data A channel detector comprising a channel estimator, a channel detector for receiving the estimated channel impulse response to determine a data extraction position, and a block extractor for extracting valid data from the superimposed equalized data according to the determined extraction position It provides a frequency domain equalizer using.

The channel detector may include a channel center detector that receives the channel impulse response and provides data extraction position information, and an extraction positioner that determines the data extraction position using the extraction position information.

The channel detector may further include a mask operator configured to receive the estimated channel impulse response and to remove noise caused by data components.

The channel detector may further include a severe channel detector that receives the estimated channel impulse response and provides channel distortion degree information.

The extraction position determiner may determine to extract data from the center of the equalized data when the channel distortion degree is equal to or less than a preset reference value.

According to an aspect of the present invention, there is provided a frequency-domain equalization method of an overlapping-storing method of overlapping and equalizing input data and extracting valid data from the equalized data, comprising: estimating a channel impulse response of the input data; And receiving a channel impulse response to determine a data extraction position, and extracting valid data from the superimposed equalized data according to the determined extraction position. .

Therefore, according to the present invention, by applying the position of removing the overlapped data differently according to the channel state in the frequency domain equalizer of the overlap-storing scheme, it is possible to increase the area that can be equalized without significantly increasing the hardware area.

Hereinafter, with reference to the accompanying drawings, preferred embodiments of the present invention that can specifically realize the above object will be described.

In addition, the terminology used in the present invention was selected as a general term that is widely used at present, but in certain cases, the term is arbitrarily selected by the applicant, and in this case, the meaning is described in detail in the description of the present invention. It is to be understood that the present invention should be understood as the meaning of terms rather than names.

5 is a block diagram showing a configuration of a frequency-domain equalizer of a superposition-storage scheme using a channel detector according to the present invention. Where a detailed description of the block is required for convenience of description, it will be described with reference to the accompanying drawings.

As shown in FIG. 5, the frequency domain equalizer according to the present invention includes a superimposed block 100, an FFT unit 200, a channel information provider 300, a complex multiplier 400, and an IFFT unit 500. And a channel estimator 700 for obtaining channel information, a channel detector 800 for determining a data extraction position using the estimated channel information, and a superimposition using the extraction position information. Block extractor 600 for extracting the extracted data is included.

The operation relationship of the frequency domain equalizer using the channel estimator configured as described above is as follows.

First, input data u (n) overlaps data by forming a data block including new data currently input through the overlapper 100 and previously input data. The overlapped data is converted into the data U (k) in the frequency domain by the FFT unit 200, and the U (k) is updated by the channel information provider 300 according to the channel state and provided by the equalization coefficient W (k And multiplied by the complex multiplier 400 to obtain data Y (k) in the equalized frequency domain.

The data Y (k) of the equalized frequency domain is inversely transformed back into the time domain through the IFFT unit 500 and input to the block extractor 600.

Meanwhile, the channel estimator 700 uses a channel impulse response (CIR) using a repetitive training sequence or a decision value of data included in the input data u (n). Estimate

The estimated channel impulse response value is input to the channel detector 800, and the channel detector 800 determines and extracts a data extraction position using the channel impulse response value. This will be described with reference to FIGS. 6 to 7 as follows.

6 is a block diagram illustrating an internal configuration of a channel detector according to an embodiment of the present invention.

As shown in FIG. 6, the channel detector 800 according to the present invention includes a mask mask 810 for removing a noise component due to a data component from input channel information, and the masked channel information. And a channel center detector 820 for receiving the data extraction position information and an extraction offset decision 830 for determining the extraction position using the extraction position information. .

을 초과하면 출력하고 그렇지 않으면 0을 출력하는 연산이다.Referring to the operation relationship of the channel detector 800 configured as described above, the channel impulse response value, which is the channel information estimated through the channel estimator 700 of FIG. 5, is input to the internal mask operator 810 of the channel detector 800. The mask operator 810 performs a mask operation to remove noise components due to data components from the estimated channel information. In the mask operation, the absolute value of the input signal is a reference value.

If exceeded, it is outputted. Otherwise, it is 0.

의 시간 영역 채널 정보 h(n)을 이용하여 다음의 수학식

,

와 같이 두 가지 기준으로 데이터 추출 위치 정보를 제공하게 된다.The channel information masked in this manner becomes an input value of the channel center detector 820. In the channel center detector 820, data sampled in symbol units

Using the time domain channel information h (n) of the following equation

,

As shown below, data extraction location information is provided based on two criteria.

과

는 위치를 맞추기 위한 상수이다. 상기 수학식 1 에서

은 채널 응답의 길이에 상관없이 중앙을 기준으로 어느 쪽이 더 많은 응답값이 있는지를 비교하기 위한 수식이며,

는 채널 응답의 길이에 따라 가중치를 달리하여 계산하는 수식이다.Where above

and

Is a constant for positioning. In Equation 1

Is a formula for comparing which of the response values is more based on the center regardless of the length of the channel response.

Is a formula calculated by varying the weight depending on the length of the channel response.

또는

는 추출 위치를 결정하기 위한 추출 위치 결정기(830)로 입력된다. 상기 추출 위치 결정기(830)에서는 상기 추출 위치 정보

또는

를 입력받아 최총 추출 위치를 결정하여 출력한다. Extraction location information calculated through Equation 1

or

Is input to the extraction positioner 830 for determining the extraction position. The extraction positioner 830, the extraction position information

or

Is inputted to determine the maximum extraction position and output.

또는

를 통해 채널 임펄스 응답이 많은 쪽의 영역을 선택하여 최총 추출 위치 P를 결정하여 출력하게 되는 것이다.That is, the extraction position determiner 830 is the extraction position information

or

By selecting the region with the higher channel impulse response, the maximum extraction position P is determined and output.

On the other hand, Figure 7 is a block diagram showing the internal configuration of a channel detector according to another embodiment of the present invention.

As shown in FIG. 7, the channel detector according to another embodiment of the present invention has a form in which a heavy channel detector 840 is added to the channel detector shown in FIG. 6.

That is, the degree of distortion of the channel is determined by the severe channel detector 840, and the extraction position determiner 830 is used as a reference for determining the extraction position of the data according to the determined result.

In general, when the channel information of the time domain is looked at in the frequency domain, it appears as shown in FIG.

That is, in the case of a channel that is severely distorted due to the strong ghost, a spectrum appears as shown in (c) of FIG. 8, and a channel having such a spectrum converges as shown in (a) or (b) of FIG. 3. (C) has a characteristic of convergence (even if there is only a casual component or an anti-casual component based on the center, as shown in (a) or (b) of FIG. 2).

Therefore, the information on the degree of distortion of the channel can be a good reference for determining the extraction position of the data.

Determination of the degree of signal distortion in the severe channel detector 840 may be determined by a response of the frequency side of the channel as H (k), and calculated by Equation 2 below.

,

는 상수이다. 상기 수학식 2에서

또는

값이 기준값(Th : Threshold) 이하인 경우 왜곡이 심한 채널이라 판단하게 된다.here

,

Is a constant. In Equation 2

or

If the value is below the threshold (Th: Threshold), it is determined that the channel is severely distorted.

또는

)와 심한 채널 검출기(840)에서 제공하는 채널 왜곡 정도 정보(

또는

)를 입력받아 최종 추출 위치 정보(P)를 다음과 같이 결정하여 출력한다.Accordingly, according to another exemplary embodiment of the present invention, the extraction position determiner 830 extracts the extraction position information from the channel center detector 820.

or

) And channel distortion degree information provided by the severe channel detector 840 (

or

) And the final extraction position information (P) is determined as follows and output.

P = A, if B> Th

P = 0, if B <= Th

Th is a threshold that determines the degree of distortion. When the B value representing the channel distortion degree is greater than the reference value as shown in Equation 3, the extraction position P is determined according to the extraction position information A value calculated by the channel center detector 820, and the B value is equal to or less than the reference value. In this case, it is determined that the degree of channel distortion is severe so as not to move the extraction position (P = 0). If the value of B is less than or equal to the reference value, it may be desirable to detect it in the center of the calculated block (see FIG. 4C).

According to the exemplary embodiment of the present invention, the position information determined by the extraction position determiner 830 is input to the block extractor 600 and used to extract valid data from the equalized and output data block.

On the other hand, the present invention is not limited to the above-described embodiment, and can be modified by those skilled in the art as can be seen from the appended claims and such modifications are within the scope of the present invention. Belongs.

The effects of the frequency domain equalization apparatus and method using the channel detector according to the present invention described above are as follows.

First, by applying different positions according to channel conditions, the frequency-equalizer of the overlap-storing frequency domain equalizer can be used to increase the area that can be equalized without significantly increasing the hardware area.

Second, there is an effect of increasing the area that can be equalized without increasing hardware speed, that is, without increasing power consumption.

Claims (9)

In the overlap-storing frequency domain equalizer for overlapping and equalizing input data and extracting valid data from the equalized data, A channel estimator for estimating a channel impulse response of the input data; A channel detector configured to receive the estimated channel impulse response and determine a valid data extraction position; And a block extractor for extracting valid data from the superimposed equalized data according to the determined extraction position. The method of claim 1, wherein the channel estimator, And a channel impulse response using an iterative training sequence included in the input data or a determined value of the data. The method of claim 1, wherein the channel detector, A channel center detector which receives the channel impulse response and provides data extraction position information; And an extraction position determiner configured to determine a data extraction position by using the extraction position information. The method of claim 3, wherein the channel detector, And a mask operator for receiving the estimated channel impulse response and removing noise due to data components. The method of claim 3, wherein the channel detector, And a severe channel detector which receives the estimated channel impulse response and provides channel distortion degree information. The method of claim 5, And the extraction position determiner determines to extract data at the center of the equalized data when the degree of distortion of the channel is equal to or less than a preset reference value. The method of claim 3, wherein The data extraction position information is a frequency domain equalizer using a channel detector, characterized in that the comparison of which the more response value based on the center irrespective of the length of the channel response. The method of claim 3, wherein The data extraction position information is a frequency domain equalizer using a channel detector, characterized in that calculated by varying the weight according to the length of the channel response. In the frequency-domain equalization method of the overlap-storing method of overlapping and equalizing input data and extracting valid data from the equalized data, Estimating a channel impulse response of the input data; Determining the valid data extraction position by receiving the estimated channel impulse response; And extracting valid data from the superimposed equalized data according to the determined extraction position.

Images