KR100281390B1 - ８-VSB Adaptive Channel Equalizer and Coefficient Update Method - Google Patents

Abstract

The present invention relates to an 8-VSB adaptive channel equalizer with a coefficient update rate and a coefficient update method. In such an 8-VSB adaptive channel equalizer, the M-tap forward filter filters the received I-channel input symbol using an activated tap according to the first tap control signal, and the first tap coefficient updater uses an error signal. Update the tap coefficient of the M-tap forward filter. Meanwhile, the N-tap feedback filter filters the equalized I-channel input symbol by using the tap activated according to the second tap control signal, and the second tap coefficient updating unit updates the tap coefficient of the N-tap feedback filter by using an error signal. do. The first subtractor generates an error signal from the difference between the equalized I-channel input symbol and the input signal of the N-tap feedback filter, and the second subtractor generates the difference between the output of the M-tap forward filter and the output of the N-tap feedback filter. Obtain The control unit stores the coefficients of the main tap of the M-tap forward filter and the first to third threshold values, and compares the error signals output from the first subtractor with the first to third threshold values, and the first and second threshold values. According to a result of comparing the tap coefficient and the main tap coefficient output from the tap coefficient updating unit, first and second tap control signals for deciding whether to activate taps of the M-tap forward filter and the N-tap feedback filter are generated. According to this configuration, the convergence speed of the equalizer is improved by adjusting the number of taps of the forward filter and the feedback filter constituting the 8-VSB adaptive channel equalizer according to the filter coefficient and the error signal value.

Description

8-VSB adaptive channel equalizer and method of updating coefficients

The present invention relates to an 8-VSB (Vestigial SideBand) adaptive channel equalizer in digital TV, and more particularly, to an 8-VSB adaptive channel equalizer and a coefficient update method with improved coefficient update rate.

Unlike conventional analog TV, digital TV converts and transmits video and audio signals to digital, so that the original signal can be received as it is without distortion of the signal due to transmission noise, and the video and audio data can be compressed and expanded. Therefore, it is possible to transmit a larger amount of data than the analog transmission method in the transmission channel of the same band, thus enabling high quality HDTV broadcasting. The advantage is that more than one program can be sent simultaneously on a channel. Modulation schemes for fully digital HDTV use 16/32 quadrature amplitude modulation (QAM) for cable broadcasting or 8/16 VSB for terrestrial broadcasting. This is because multi-level modulation is inevitable in order to transmit a large amount of data generated after encoding through the 6MHz band, which is an existing NTSC transmission channel. The QAM method is one of very excellent transmission efficiency. Since the signal is represented in two dimensions, the distance between signals is relatively long, which increases the transmission efficiency. However, since the signal itself is represented by a complex number, the hardware cost is relatively increased. There is a disadvantage. On the other hand, in contrast to the QAM method, the VSB method has a one-dimensional constellation, so hardware for processing data is simple, but the distance between signals is small, so that the symbol error rate is large and the modulation / demodulation system is relatively complicated.

On the other hand, the signal transmitted from the transmitting end is caused various distortions through the transmission channel. Factors that cause distortion include deformation by Gaussian thermal noise, addition or multiplication noise due to fading, frequency variation, nonlinearity, and time dispersion. Such distortion appears in the picture quality due to distortion in the existing analog TV system, but in the digital transmission system, a bit detection error occurs at the receiver, which may result in impossible or unexpected data recovery. In particular, multipath due to time delay and phase change of a transmission signal causes severe intersymbol interference, which is a major cause of bit detection error. The technique of reducing the bit detection error at the receiver by compensating for distortion caused by the non-ideal transmission channel is called channel equalization. However, since the channel is variable by various factors such as the location, distance, terrain, buildings, and weather of the transceiver, an equalization technique that can be adaptively substituted for the variable channel is required. Such a technique is called adaptive channel equalization. To this end, by adjusting the tap coefficient so that the transfer function of the equalizer is the inverse of the transfer function of a given channel, the linear distortion in the channel is equalized to eliminate the channel distortion effect called interference between adjacent symbols. The tap coefficient is continuously updated to equalize channel characteristics at a given time.

Figure 1 shows a typical 8-VSB adaptive channel equalizer, which is an example of a real adaptive equalizer consisting of 256 taps. The 8-VSB adaptive channel equalizer uses the coefficients of the DC offset remover 11, the 64 tap feed forward filter 12, and the 64 tap feed forward filter 12 to remove the DC offset by the circuit offset or the pilot. Selector 15, 192 taps for determining one of a training signal or a direct determination signal as an input signal of the first filter coefficient updater 13, 192 tap feedback filter 14, and 192 tap feedback filter 14 to be updated. And a second filter coefficient updater 16 for updating the filter coefficients of the feedback filter 14, and first and second subtractors 17 and 18.

In the adaptive channel equalizer of FIG. 1, a Least Mean Square (LMS) algorithm is applied for coefficient update, which is performed by the following equation.

C _{k + 1} = C _k + με _k I _k

Where C _{k + 1} is a new set of coefficients, such as {c _{k + 1, -63} , c _{k + 1, -62} , ..., c _{k + 1,0} , c _{k + 1,1} , c _{k + 1,2} , ..., c _{k + 1,192} }, where C _k is the previous set of coefficients, where {c _{k, -63} , c _{k, -62} , ..., c _{k, 0} , c _{k, 1} , c _{k, 2} , ..., c _{k, 192} }, and I _k is an input sequence, {I _-63 , I _-62 , ..., I ₀ , I ₁ , I _{2 ,.} ..., I ₁₉₂ }, and ε _k is an error sequence, expressed as {e _-63 , e _-62 , ..., e ₀ , e ₁ , e ₂ , ..., e ₁₉₂ }. In Equation 1, e _i is a value obtained by subtracting an output value from an input value of a signal determiner at an instant of i, also called an error value.

In the conventional adaptive channel equalizer, in order to improve the convergence speed, the step size, i. In other words, in the initial stage, before the adaptive channel equalizer is still converged, fast convergence is achieved by updating the coefficients of the equalizer using a large step size (μ), and then small to reduce the effects of noise and the like. The method of using the step size (mu) of the value is used.

However, the adaptive channel equalizer for terrestrial broadcasting requires a great deal of computation to update the coefficients as in Equation 1 above. According to Equation 1, since the amount of calculation is proportional to the number of taps of the equalizer, the amount of calculation increases as the number of taps increases. Therefore, the coefficient update cannot be performed for every data symbol, but for several data symbols, for example, the coefficient update must be performed once for every 2,4,8,16,32,64 or 128 data symbols. Even using (μ), there is a problem that the time taken for the equalizer to converge becomes long.

Therefore, the present invention has been proposed to solve the above-described problem, and in order to quickly update coefficients by using only some of the taps around the main tab instead of all the taps of the equalizer at the stage before the equalizer converges. The purpose of the present invention is to provide an 8-VSB adaptive channel equalizer and its coefficient update method.

In order to achieve the above object, the 8-VSB adaptive equalizer includes: an M-tap forward filter for filtering a received I-channel input symbol using an activated tap according to a first tap control signal; A first tap coefficient updating unit updating the tap coefficient of the M tap forward filter by using an error signal; An N-tap feedback filter for filtering the equalized I-channel input symbols using taps activated according to the second tap control signal; A second tap coefficient updating unit updating the tap coefficient of the N tap feedback filter by using the error signal; A first subtractor for generating an error signal from a difference value between the equalized I-channel input symbol and an input signal of the N-tap feedback filter; A second subtractor for obtaining a difference value between an output of the M-tap forward filter and an output of the N-tap feedback filter; And a coefficient of the main tap of the M-tap forward filter and first to third threshold values, wherein a comparison result of the error signal output from the first subtractor and the first to third threshold values is used. And generating first and second tap control signals for determining whether taps of the M-tap forward filter and the N-tap feedback filter are activated according to a result of comparing the tap coefficient and the main tap coefficient output from the second tap coefficient updater. It characterized in that it comprises a control unit to make.

In order to achieve the above object, the coefficient updating method of the 8-VSB adaptive equalizer according to the present invention includes activating some taps of the M-tap forward filter, and then counting the coefficients of the main tap and the number of numbers of the remaining taps, and errors. A first tap activation step of activating all taps of the M-tap forward filter according to a result of comparing the signal value with a predetermined first threshold value; And activating some taps of the N-tap feedback filter according to the result of comparing the coefficients of the main tap and the coefficients of the remaining taps of the M-tap forward filter, and the error signal value and the predetermined second threshold value. And a second tap activation step of activating all taps of the N-tap feedback filter according to the result of comparing the coefficient of the main tap with the coefficient magnitude of the remaining tap and the error signal value and the predetermined third threshold value.

1 is a block diagram showing a typical 8-VSB adaptive channel equalizer;

2 is a block diagram illustrating an 8-VSB adaptive channel equalizer according to the present invention;

3 is a flowchart illustrating a coefficient updating method of an 8-VSB adaptive channel equalizer according to the present invention; and

4A and 4B illustrate an FFE tap and a DFE tap that are activated over time, respectively.

* Explanation of symbols for main parts of the drawings

21 ... M-tap forward filter 22 ... First tap coefficient update unit

23 ... N-tap feedback filter 24 ... Input signal selector

24-1 ... training signal generator 24-2 ... signal determiner

24-3 ... Multiplexer 25 ... Second tap coefficient update unit

26,27 ... Subtractor 28 ... Controls

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 2 is a block diagram illustrating an 8-VSB adaptive channel equalizer according to the present invention, and includes a first tap coefficient updating unit for updating tap coefficients of the M tap forward filter 21 (FFE) and the M tap forward filter 21 ( 22) an input signal selector 24 for selecting input signals of the N tap feedback filter 23 (DFE) and the N tap feedback filter 23, and a second tap for updating the tap coefficient of the N tap feedback filter 23; The coefficient updater 25, the first subtractor 26 for generating an error signal, and the second subtractor 27 for generating a difference between the output of the M-tap forward filter 21 and the output of the N-tap feedback filter 23. And a tap control signal for controlling taps of the M-tap forward filter 21 and the N-tap feedback filter 23 by inputting tap coefficients output from the first and second tap coefficient update units 22 and 25. It consists of a control unit 28 to make. Here, the tap number M of the FFE 21 and the tap number N of the DFE 23 may be 64 and 192, respectively, as shown in FIG. 1, or may have different values according to a user's selection.

3 is a flowchart illustrating a method of updating a coefficient of an 8-VSB adaptive channel equalizer according to the present invention, which is performed by the controller 28 of FIG. In the coefficient updating method of the 8-VSB adaptive channel equalizer according to the present invention, after activating some taps of the M-tap forward filter 21, the magnitudes of the main tap coefficients and the remaining tap coefficients, the error signal values, and the predetermined first First tap activation steps 31 to 33 for activating all taps of the M-tap forward filter 21 according to the comparison result of the threshold value, the magnitudes of the main tap coefficients and the remaining tap coefficients, the error signal value, and a predetermined value. After activating some taps of the N-tap feedback filter 23 according to the comparison result of the two thresholds, N according to the magnitude of the main tap 곗 and the remaining tap coefficients again, and the comparison result of the error signal value and the predetermined third threshold value. The second tap activation step 34 to 37 activates all taps of the tap feedback filter 23.

Next, the operation of the 8-VSB adaptive channel equalizer according to the present invention will be described in detail with reference to FIGS. 2 and 3.

In FIG. 2, the M-tap forward filter 21 is for filtering an I-channel input symbol (10.76 MS / sec) among the received 8-VSB signals. First, only some taps around the main tap are activated. The controller determines whether to activate the remaining taps according to the first tap control signal output from the controller 28. Here, the main tap is arbitrarily determined or automatically changed by the user according to the characteristics of the signal input to the M-tap forward filter 21. The first tap coefficient updating unit 22 updates the tap coefficient according to the error signal output from the first subtractor 26 and supplies the tap coefficient to the M-tap forward filter 21.

The N-tap feedback filter 23 is a decision feedback filter, which is a training signal generator 24 of the input signal selector 24 at the beginning of reception in which the multiplexer 24-3 has no knowledge of channel characteristics. -1) transmits the training signal outputted by the controller, and determines the tap coefficients of the N-tap feedback filter 23 so that the distortion characteristics of the channel are canceled during this period, and is determined by the signal determiner 24-2 at the end of this period. In decision-directed mode, normal data transfer is achieved. The second tap coefficient updating unit 25 updates the tap coefficient according to the error signal output from the first subtractor 26 and supplies the tap coefficient to the N tap feedback filter 23.

The first subtractor 26 calculates a difference between the equalized I-channel input symbol output from the second subtractor 27 and the signal output from the input signal selector 24 and uses the first and second tap coefficients as error signals. It supplies to the update part 22,25 and the control part 28. The second subtractor 27 obtains the difference between the output of the M-tap forward filter 21 and the output of the N-tap feedback filter 23 and outputs the equalized I-channel input symbol.

The controller 28 stores the coefficients of the main tap of the M-tap forward filter 21 and the first to third threshold values, and the error signal and the first to third threshold values output from the first subtractor 26. The taps of the M-tap forward filter 21 and the N-tap feedback filter 23 according to the comparison result and the comparison result of the tap coefficients and the main tap coefficients output from the first and second tap coefficient updating units 22 and 25. First and second tap control signals for determining whether to activate are generated and supplied to the M-tap forward filter 21 and the N-tap feedback filter 23, respectively.

3 is a diagram illustrating a tap activation process of the adaptive channel equalizer of FIG. 2, which is performed by comparing magnitudes of an error signal and a tap coefficient in an MCU which is a controller (28 of FIG. 2).

FIG. 4 shows the FFE 21 and DFE 23 taps activated over time by the coefficient updating process of FIG. 3, wherein the tabs shown in black are the active taps used in the coefficient update, and the white taps are the coefficients. Deactivate tab not used for update.

Next, the tap activation process of FIG. 3 will be described with reference to FIG. 4.

In the initial stage of adaptive equalization, that is, in the thirty-first stage, only a part of taps are activated around the main tap of the M-tap forward filter 21 ((1) and (2) of FIG. 4), and in the thirty-second stage, the main tap coefficient Compare the size of with the size of the activated tap coefficient. At this time, the coefficients of the remaining taps which are not activated are all fixed to 0 until they are activated. In this state, if any of the activated taps have a tap larger than the main tap coefficient, the equalizer has not yet converged and waits for convergence. If the main tap coefficient is larger than all activated tap coefficients, the equalizer is converged. From this time, the magnitude of the error signal output from the first subtractor 26 is compared with the first threshold value in addition to the comparison of the coefficient magnitudes in step 32. . If the error value is greater than the first threshold, wait until the equalizer is more stable. If the error value is smaller than the first threshold value, the equalizer is converged from the FFE 21 taps in which only some taps are activated, and thus activates all the other FFE 21 taps (step 33; FIG. 4 (3)).

In a thirty-fourth step, the magnitude of the main tap coefficient of the FFE 21 is compared with all activated tap coefficients; and if the main tap coefficient is larger than all the remaining tap coefficients, the magnitude of the error signal output from the first subtractor 26 is reduced. The second threshold is smaller than the one threshold. If the magnitude of the error signal is smaller than the second threshold, only a part of the tap of the DFE 23 is activated (step 35; FIG. 4 (4)). In the 36th step, the magnitude of the main tap coefficient of the FFE 21 is compared with all activated tap coefficients. When the main tap coefficient is larger than all the remaining tap coefficients, the magnitude of the error signal output from the first subtractor 26 is reduced. Compared to the third threshold smaller than the second threshold. If the magnitude of the error signal is smaller than the third threshold, all taps of the DFE 23 are activated (step 37; FIG. 4 (5)). Here, the first to third threshold values used in the above steps are experimentally determined values.

According to the above-described configuration, before the equalizer converges, only some taps are used around the main tap of the forward filter without using all the taps of the forward filter and the feedback filter. By activating the tap, the convergence speed of the equalizer can be improved, thereby improving the equalizer's adaptability to rapidly changing dynamic ghosts in time.

Claims (4)

An M-tap forward filter 21 for filtering the received I-channel input symbol using an activated tap according to the first tap control signal; A first tap coefficient updating unit 22 for updating the tap coefficient of the M tap forward filter 21 by using an error signal; An N-tap feedback filter 23 for filtering the equalized I-channel input symbols using taps activated according to the second tap control signal; A second tap coefficient updating unit 25 for updating the tap coefficient of the N-tap feedback filter 23 by using the error signal; A first subtractor (26) for generating the error signal from a difference value between the equalized I-channel input symbol and the input signal of the N-tap feedback filter (23); A second subtractor (27) for obtaining a difference value between the output of the M-tap forward filter (21) and the output of the N-tap feedback filter (23); And Coefficients of the main tap of the M-tap forward filter 21 and first to third threshold values are stored, and a result of comparing the error signal output from the first subtractor 26 with the first to third threshold values. And taps of the M-tap forward filter 21 and the N-tap feedback filter 23 according to a comparison result of the tap coefficients and the main tap coefficients output from the first and second tap coefficient update units 22 and 24. And a control unit (28) for generating first and second tap control signals for determining whether to activate each. After activating some taps around the main tap of the M-tap forward filter 21, the magnitude of the coefficients of the main tap and the number of numbers of the remaining taps, and the input signal of the equalized I-channel input symbol and the N-tap feedback filter 23 First tap activation steps 31 to 33 activating all taps of the M-tap forward filter 21 according to a comparison result of an error signal value generated from a difference value with a predetermined first threshold value; And The N-tap feedback filter 23 according to the result of comparing the coefficient of the main tap with the coefficient of the remaining tap of the M-tap forward filter 21 and the error signal value and the predetermined second threshold smaller than the first threshold value. After activating some taps, the coefficients of the main tap and the remaining tap coefficients of the M-tap forward filter 21 and the comparison result of the error signal value and the predetermined third threshold smaller than the second threshold value are determined. And a second tap activation step (34-37) for activating all taps of the N-tap feedback filter (23). The method of claim 2, wherein the first tap activation step 31 to 33 is performed. Activating (31) a part of the taps around the main tap of the M-tap forward filter (21); Comparing the magnitude of the main tap coefficient of the M-tap forward filter 21 with the remaining tap coefficients, and comparing the magnitude of the error signal value with the first threshold value when the main tap coefficient is larger than all the remaining tap coefficients; And Activating (33) all taps of the M-tap forward filter (21) if the magnitude of the error signal value is less than a first threshold. . The method of claim 3, wherein the second tap activation step (34 ~ 37) Comparing the magnitude of the main tap coefficient of the M-tap forward filter 21 with the remaining tap coefficients, and comparing the magnitude of the error signal value with the second threshold value when the main tap coefficient is larger than all the remaining tap coefficients; Activating (35) some taps of the N-tap feedback filter (23) if the magnitude of the error signal value is less than a second threshold value; Comparing the magnitude of the main tap coefficient of the M tap forward filter 21 with the remaining tap coefficients, and comparing the magnitude of the error signal value with a third threshold value when the main tap coefficient is larger than all the remaining tap coefficients; And Activating all the taps of the N-tap feedback filter 23 if the magnitude of the error signal value is smaller than a third threshold, and updating the coefficients of the 8-VSB adaptive channel equalizer. .